init model

README.md

@@ -0,0 +1,148 @@
+---
+license: apache-2.0
+datasets:
+- COCO
+metrics:
+- mAP
+language:
+- en
+tags:
+- RyzenAI
+- object-detection
+- vision
+- YOLO
+- Pytorch
+---
+
+# YOLOv3 model trained on COCO
+
+YOLOv3 is trained on COCO object detection (118k annotated images) at resolution 416x416. It was released in https://github.com/ultralytics/yolov3/tree/v8.
+
+We develop a modified version that could be supported by [AMD Ryzen AI](https://ryzenai.docs.amd.com).
+
+
+## Model description
+
+YOLOv3 🚀 is the world's most loved vision AI, representing Ultralytics open-source research into future vision AI methods, incorporating lessons learned and best practices evolved over thousands of hours of research and development.
+
+
+## Intended uses & limitations
+
+You can use the raw model for object detection. See the [model hub](https://huggingface.co/models?search=amd/yolov3) to look for all available YOLOv3 models.
+
+
+## How to use
+
+### Installation
+
+   Follow [Ryzen AI Installation](https://ryzenai.docs.amd.com/en/latest/inst.html) to prepare the environment for Ryzen AI.
+   Run the following script to install pre-requisites for this model.
+   ```bash
+   pip install -r requirements.txt 
+   ```
+
+
+### Data Preparation (optional: for accuracy evaluation)
+
+The dataset MSCOCO2017 contains 118287 images for training and 5000 images for validation.
+
+1. Download COCO dataset
+2. Run general_json2yolo.py to generate the labels folder and val2017.txt
+  ```sh
+  python general_json2yolo.py
+  ```
+Finally, COCO dataset should look like this:
+```plain
++ coco/
+    + annotations/
+        + instance_val2017.json
+        + ...
+    + images/
+        + val2017/
+          + 000000000139.jpg
+          + 000000000285.jpg
+          + ...   
+    + labels/
+        + val2017/
+          + 000000000139.txt
+          + 000000000285.txt
+          + ...      
+    + val2017.txt
+```
+
+
+
+### Test & Evaluation
+
+ - Code snippet from [`onnx_inference.py`](onnx_inference.py) on how to use
+```python
+    onnx_path = "yolov3-8.onnx"
+    onnx_model = onnxruntime.InferenceSession(
+        onnx_path, providers=providers, provider_options=provider_options)
+
+    path = opt.img
+    new_path = os.path.join(opt.out, "demo_infer.jpg")
+
+    conf_thres, iou_thres, classes, agnostic_nms, max_det = 0.25, \
+        0.45, None, False, 1000
+
+    img0 = cv2.imread(path)
+    img = pre_process(img0)
+    onnx_input = {onnx_model.get_inputs()[0].name: img}
+    onnx_output = onnx_model.run(None, onnx_input)
+    onnx_output = post_process(onnx_output)
+
+    pred = non_max_suppression(
+        onnx_output[0],
+        conf_thres,
+        iou_thres,
+        multi_label=False,
+        classes=classes,
+        agnostic=agnostic_nms)
+
+    colors = [[random.randint(0, 255) for _ in range(3)]
+              for _ in range(len(names))]
+    det = pred[0]
+    im0 = img0.copy()
+
+    if len(det):
+        # Rescale boxes from imgsz to im0 size
+        det[:, :4] = scale_coords(img.shape[2:], det[:, :4], im0.shape).round()
+
+        # Write results
+        for *xyxy, conf, cls in reversed(det):
+            label = '%s %.2f' % (names[int(cls)], conf)
+            plot_one_box(xyxy, im0, label=label, color=colors[int(cls)])
+
+    # Stream results
+    cv2.imwrite(new_path, im0)
+```
+
+ - Run inference for a single image
+  ```sh
+  python onnx_inference.py --img INPUT_IMG_PATH --out OUTPUT_DIR --ipu --provider_config Path\To\vaip_config.json
+  ```
+  *Note: __vaip_config.json__ is located at the setup package of Ryzen AI (refer to [Installation](#installation))*
+
+ - Test accuracy of the quantized model
+  ```sh
+  python onnx_test.py --ipu --provider_config Path\To\vaip_config.json
+  ```
+
+### Performance
+
+|Metric |Accuracy on IPU|
+| :----:  | :----: |
+|AP\@0.50:0.95|0.389|
+
+
+```bibtex
+@misc{redmon2018yolov3,
+      title={YOLOv3: An Incremental Improvement}, 
+      author={Joseph Redmon and Ali Farhadi},
+      year={2018},
+      eprint={1804.02767},
+      archivePrefix={arXiv},
+      primaryClass={cs.CV}
+}
+```

coco.names

@@ -0,0 +1,80 @@
+person
+bicycle
+car
+motorcycle
+airplane
+bus
+train
+truck
+boat
+traffic light
+fire hydrant
+stop sign
+parking meter
+bench
+bird
+cat
+dog
+horse
+sheep
+cow
+elephant
+bear
+zebra
+giraffe
+backpack
+umbrella
+handbag
+tie
+suitcase
+frisbee
+skis
+snowboard
+sports ball
+kite
+baseball bat
+baseball glove
+skateboard
+surfboard
+tennis racket
+bottle
+wine glass
+cup
+fork
+knife
+spoon
+bowl
+banana
+apple
+sandwich
+orange
+broccoli
+carrot
+hot dog
+pizza
+donut
+cake
+chair
+couch
+potted plant
+bed
+dining table
+toilet
+tv
+laptop
+mouse
+remote
+keyboard
+cell phone
+microwave
+oven
+toaster
+sink
+refrigerator
+book
+clock
+vase
+scissors
+teddy bear
+hair drier
+toothbrush

coco2017.data

@@ -0,0 +1,4 @@
+classes=80
+train=coco/train2017.txt
+valid=coco/val2017.txt
+names=coco.names

general_json2yolo.py

@@ -0,0 +1,179 @@
+import numpy as np
+from tqdm import tqdm
+from pathlib import Path
+import json
+from collections import defaultdict
+import sys
+import pathlib
+
+CURRENT_DIR = pathlib.Path(__file__).parent
+sys.path.append(str(CURRENT_DIR))
+
+
+def make_dirs(path='coco'):
+    # Create folders
+    path = Path(path)
+    for p in [path / 'labels']:
+        p.mkdir(parents=True, exist_ok=True)  # make dir
+    return path
+
+
+def coco91_to_coco80_class():  # converts 80-index (val2014) to 91-index (paper)
+    # https://tech.amikelive.com/node-718/what-object-categories-labels-are-in-coco-dataset/
+    x = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, None, 11, 12, 13, 14, 15, 16, 17,
+         18, 19, 20, 21, 22, 23, None, 24, 25, None, None, 26, 27, 28, 29,
+         30, 31, 32, 33, 34, 35, 36, 37, 38, 39, None, 40, 41, 42, 43, 44,
+         45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, None,
+         60, None, None, 61, None, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71,
+         72, None, 73, 74, 75, 76, 77, 78, 79, None]
+    return x
+
+
+def convert_coco_json(
+        json_dir='coco/annotations/',
+        use_segments=False,
+        cls91to80=False):
+    save_dir = make_dirs()  # output directory
+    coco80 = coco91_to_coco80_class()
+    """Convert raw COCO dataset to YOLO style
+    """
+
+    # Import json
+    for json_file in sorted(Path(json_dir).resolve().glob('instances_val2017.json')):
+        fn = Path(save_dir) / 'labels' / \
+             json_file.stem.replace('instances_', '')  # folder name
+        fn.mkdir()
+        with open(json_file) as f:
+            data = json.load(f)
+
+        # Create image dict
+        images = {'%g' % x['id']: x for x in data['images']}
+        # Create image-annotations dict
+        imgToAnns = defaultdict(list)
+        for ann in data['annotations']:
+            imgToAnns[ann['image_id']].append(ann)
+
+        txt_file = open(Path(save_dir / 'val2017').
+                        with_suffix('.txt'), 'a')
+        # Write labels file
+        for img_id, anns in tqdm(
+                imgToAnns.items(), desc=f'Annotations {json_file}'):
+            img = images['%g' % img_id]
+            h, w, f = img['height'], img['width'], img['file_name']
+            bboxes = []
+            segments = []
+
+            txt_file.write(
+                './images/' + '/'.
+                join(img['coco_url'].split('/')[-2:]) + '\n')
+            for ann in anns:
+                if ann['iscrowd']:
+                    continue
+                # The COCO box format is
+                # [top left x, top left y, width,
+                # height]
+                box = np.array(ann['bbox'], dtype=np.float64)
+                box[:2] += box[2:] / 2  # xy top-left corner to center
+                box[[0, 2]] /= w  # normalize x
+                box[[1, 3]] /= h  # normalize y
+                if box[2] <= 0 or box[3] <= 0:  # if w <= 0 and h <= 0
+                    continue
+                cls = coco80[ann['category_id'] - 1] \
+                    if cls91to80 else ann['category_id'] - 1  # class
+                box = [cls] + box.tolist()
+                if box not in bboxes:
+                    bboxes.append(box)
+                # Segments
+                if use_segments:
+                    if len(ann['segmentation']) > 1:
+                        s = merge_multi_segment(ann['segmentation'])
+                        s = (np.concatenate(s, axis=0) /
+                             np.array([w, h])).reshape(-1).tolist()
+                    else:
+                        s = [j for i in ann['segmentation']
+                             for j in i]  # all segments concatenated
+                        s = (np.array(s).reshape(-1, 2) /
+                             np.array([w, h])).reshape(-1).tolist()
+                    s = [cls] + s
+                    if s not in segments:
+                        segments.append(s)
+
+            # Write
+            with open((fn / f).with_suffix('.txt'), 'a') as file:
+                for i in range(len(bboxes)):
+                    # cls, box or segments
+                    line = *(segments[i] if
+                             use_segments else bboxes[i]),
+                    file.write(('%g ' * len(line)).
+                               rstrip() % line + '\n')
+        txt_file.close()
+
+
+def min_index(arr1, arr2):
+    """Find a pair of indexes with the shortest distance.
+    Args:
+        arr1: (N, 2).
+        arr2: (M, 2).
+    Return:
+        a pair of indexes(tuple).
+    """
+    dis = ((arr1[:, None, :] - arr2[None, :, :]) ** 2).sum(-1)
+    return np.unravel_index(np.argmin(dis, axis=None), dis.shape)
+
+
+def merge_multi_segment(segments):
+    """Merge multi segments to one list.
+    Find the coordinates with min distance between each segment,
+    then connect these coordinates with one thin line to merge all
+    segments into one.
+
+    Args:
+        segments(List(List)): original
+            segmentations in coco's json file.
+            like [segmentation1, segmentation2,...],
+            each segmentation is a list of coordinates.
+    """
+    s = []
+    segments = [np.array(i).reshape(-1, 2) for i in segments]
+    idx_list = [[] for _ in range(len(segments))]
+
+    # record the indexes with min distance between each segment
+    for i in range(1, len(segments)):
+        idx1, idx2 = min_index(segments[i - 1], segments[i])
+        idx_list[i - 1].append(idx1)
+        idx_list[i].append(idx2)
+
+    # use two round to connect all the segments
+    for k in range(2):
+        # forward connection
+        if k == 0:
+            for i, idx in enumerate(idx_list):
+                # middle segments have two indexes
+                # reverse the index of middle segments
+                if len(idx) == 2 and idx[0] > idx[1]:
+                    idx = idx[::-1]
+                    segments[i] = segments[i][::-1, :]
+
+                segments[i] = np.roll(segments[i], -idx[0], axis=0)
+                segments[i] = np.concatenate([segments[i],
+                                              segments[i][:1]])
+                # deal with the first segment and the last one
+                if i in [0, len(idx_list) - 1]:
+                    s.append(segments[i])
+                else:
+                    idx = [0, idx[1] - idx[0]]
+                    s.append(segments[i][idx[0]:idx[1] + 1])
+
+        else:
+            for i in range(len(idx_list) - 1, -1, -1):
+                if i not in [0, len(idx_list) - 1]:
+                    idx = idx_list[i]
+                    nidx = abs(idx[1] - idx[0])
+                    s.append(segments[i][nidx:])
+    return s
+
+
+if __name__ == '__main__':
+    convert_coco_json('coco/annotations',
+                      use_segments=False,
+                      cls91to80=True)

onnx_inference.py

@@ -0,0 +1,151 @@
+import onnxruntime
+import argparse
+import os
+from utils import *
+
+
+def pre_process(img):
+    """
+    Preprocessing part of YOLOv3 for scaling and padding image as input to the network.
+    Args:
+        img (numpy.ndarray): H x W x C, image read with OpenCV
+    Returns:
+        padded_img (numpy.ndarray): preprocessed image to be fed to the network
+    """
+    img = letterbox(img, auto=False)[0]
+    # Convert
+    img = img.transpose((2, 0, 1))[::-1]  # HWC to CHW, BGR to RGB
+    img = np.ascontiguousarray(img)
+    img = img.astype("float32")
+    img = img / 255.0
+    img = img[np.newaxis, :]
+    return img
+
+
+def post_process(x, conf_thres=0.1, iou_thres=0.6, multi_label=True,
+                 classes=None, agnostic=False):
+    """
+    Post-processing part of YOLOv3 for generating final results from outputs of the network.
+    Returns:
+        pred (torch.tensor): n x 6, dets[:,:4] -> boxes, dets[:,4] -> scores, dets[:,5] -> class indices
+    """
+    stride = [32, 16, 8]
+    anchors = [[10, 13, 16, 30, 33, 23],
+               [30, 61, 62, 45, 59, 119],
+               [116, 90, 156, 198, 373, 326]]
+    temp = [13, 26, 52]
+    res = []
+
+    def create_grids(ng=(13, 13)):
+        nx, ny = ng  # x and y grid size
+        ng = torch.tensor(ng, dtype=torch.float)
+
+        # build xy offsets
+        yv, xv = torch.meshgrid([torch.arange(ny), torch.arange(nx)])
+        grid = torch.stack((xv, yv), 2).view((1, 1, ny, nx, 2)).float()
+
+        return grid
+
+    for i in range(3):
+        out = torch.from_numpy(x[i])
+
+        bs, _, ny, nx = out.shape  # bs, 255, 13, 13
+
+        anchor = torch.Tensor(anchors[2 - i]).reshape(3, 2)
+        anchor_vec = anchor / stride[i]
+        anchor_wh = anchor_vec.view(1, 3, 1, 1, 2)
+
+        grid = create_grids((nx, ny))
+
+        out = out.view(
+            bs, 3, 85, temp[i], temp[i]).permute(
+            0, 1, 3, 4, 2).contiguous()  # prediction
+
+        io = out.clone()
+
+        io[..., :2] = torch.sigmoid(io[..., :2]) + grid
+        io[..., 2:4] = torch.exp(io[..., 2:4]) * anchor_wh
+        io[..., :4] *= stride[i]
+        torch.sigmoid_(io[..., 4:])
+
+        res.append(io.view(bs, -1, 85))
+
+    pred = non_max_suppression(torch.cat(res, 1), conf_thres,
+                               iou_thres, multi_label=multi_label,
+                               classes=classes, agnostic=agnostic)
+
+    return pred
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(
+        prog='One image inference of onnx model')
+    parser.add_argument(
+        '--img',
+        type=str,
+        help='Path of input image')
+    parser.add_argument(
+        '--out',
+        type=str,
+        default='.',
+        help='Path of out put image')
+    parser.add_argument(
+        "--ipu",
+        action="store_true",
+        help="Use IPU for inference.")
+    parser.add_argument(
+        "--provider_config",
+        type=str,
+        default="vaip_config.json",
+        help="Path of the config file for seting provider_options.")
+    parser.add_argument(
+        "--onnx_path",
+        type=str,
+        default="yolov3-8.onnx",
+        help="Path of the onnx model.")
+
+    opt = parser.parse_args()
+    with open('coco.names', 'r') as f:
+        names = f.read()
+
+    if opt.ipu:
+        providers = ["VitisAIExecutionProvider"]
+        provider_options = [{"config_file": opt.provider_config}]
+    else:
+        providers = ['CUDAExecutionProvider', 'CPUExecutionProvider']
+        provider_options = None
+
+    onnx_path = opt.onnx_path
+    onnx_model = onnxruntime.InferenceSession(
+        onnx_path, providers=providers, provider_options=provider_options)
+
+    path = opt.img
+    new_path = os.path.join(opt.out, "demo_infer.jpg")
+
+    conf_thres, iou_thres, classes, agnostic_nms, max_det = 0.25, \
+        0.45, None, False, 1000
+
+    img0 = cv2.imread(path)
+    img = pre_process(img0)
+    onnx_input = {onnx_model.get_inputs()[0].name: img}
+    onnx_output = onnx_model.run(None, onnx_input)
+    pred = post_process(onnx_output, conf_thres,
+                        iou_thres, multi_label=False,
+                        classes=classes, agnostic=agnostic_nms)
+
+    colors = [[random.randint(0, 255) for _ in range(3)]
+              for _ in range(len(names))]
+    det = pred[0]
+    im0 = img0.copy()
+
+    if len(det):
+        # Rescale boxes from imgsz to im0 size
+        det[:, :4] = scale_coords(img.shape[2:], det[:, :4], im0.shape).round()
+
+        # Write results
+        for *xyxy, conf, cls in reversed(det):
+            label = '%s %.2f' % (names[int(cls)], conf)
+            plot_one_box(xyxy, im0, label=label, color=colors[int(cls)])
+
+    # Stream results
+    cv2.imwrite(new_path, im0)

onnx_test.py

@@ -0,0 +1,1247 @@
+import argparse
+import json
+import os
+from pathlib import Path
+from tqdm import tqdm
+import glob
+import math
+from PIL import ExifTags, Image
+import shutil
+from torch.utils.data import DataLoader
+from torch.utils.data import Dataset
+from utils import *
+import onnxruntime
+import matplotlib.pyplot as plt
+
+for orientation in ExifTags.TAGS.keys():
+    if ExifTags.TAGS[orientation] == 'Orientation':
+        break
+
+
+def create_folder(path='./new_folder'):
+    # Create folder
+    if os.path.exists(path):
+        shutil.rmtree(path)  # delete output folder
+    os.makedirs(path)  # make new output folder
+
+
+def exif_size(img):
+    # Returns exif-corrected PIL size
+    s = img.size  # (width, height)
+    try:
+        rotation = dict(img._getexif().items())[orientation]
+        if rotation == 6:  # rotation 270
+            s = (s[1], s[0])
+        elif rotation == 8:  # rotation 90
+            s = (s[1], s[0])
+    except BaseException:
+        pass
+
+    return s
+
+
+def ap_per_class(tp, conf, pred_cls, target_cls):
+    """ Compute the average precision, given the recall and precision curves.
+    Source: https://github.com/rafaelpadilla/Object-Detection-Metrics.
+    # Arguments
+        tp:    True positives (nparray, nx1 or nx10).
+        conf:  Objectness value from 0-1 (nparray).
+        pred_cls: Predicted object classes (nparray).
+        target_cls: True object classes (nparray).
+    # Returns
+        The average precision as computed in py-faster-rcnn.
+    """
+
+    # Sort by objectness
+    i = np.argsort(-conf)
+    tp, conf, pred_cls = tp[i], conf[i], pred_cls[i]
+
+    # Find unique classes
+    unique_classes = np.unique(target_cls)
+
+    # Create Precision-Recall curve and compute AP for each class
+    pr_score = 0.1
+    # score to evaluate P and R
+    # https://github.com/ultralytics/yolov3/issues/898
+    # number class, number iou thresholds (i.e. 10 for mAP0.5...0.95)
+    s = [unique_classes.shape[0], tp.shape[1]]
+    ap, p, r = np.zeros(s), np.zeros(s), np.zeros(s)
+    for ci, c in enumerate(unique_classes):
+        i = pred_cls == c
+        n_gt = (target_cls == c).sum()  # Number of ground truth objects
+        n_p = i.sum()  # Number of predicted objects
+
+        if n_p == 0 or n_gt == 0:
+            continue
+        else:
+            # Accumulate FPs and TPs
+            fpc = (1 - tp[i]).cumsum(0)
+            tpc = tp[i].cumsum(0)
+
+            # Recall
+            recall = tpc / (n_gt + 1e-16)  # recall curve
+            # r at pr_score, negative x, xp because xp decreases
+            r[ci] = np.interp(-pr_score, -conf[i], recall[:, 0])
+
+            # Precision
+            precision = tpc / (tpc + fpc)  # precision curve
+            p[ci] = np.interp(-pr_score, -conf[i],
+                              precision[:, 0])  # p at pr_score
+
+            # AP from recall-precision curve
+            for j in range(tp.shape[1]):
+                ap[ci, j] = compute_ap(recall[:, j], precision[:, j])
+
+            # Plot
+            # fig, ax = plt.subplots(1, 1, figsize=(5, 5))
+            # ax.plot(recall, precision)
+            # ax.set_xlabel('Recall')
+            # ax.set_ylabel('Precision')
+            # ax.set_xlim(0, 1.01)
+            # ax.set_ylim(0, 1.01)
+            # fig.tight_layout()
+            # fig.savefig('PR_curve.png', dpi=300)
+
+    # Compute F1 score (harmonic mean of precision and recall)
+    f1 = 2 * p * r / (p + r + 1e-16)
+
+    return p, r, ap, f1, unique_classes.astype('int32')
+
+
+def time_synchronized():
+    torch.cuda.synchronize() if torch.cuda.is_available() else None
+    return time.time()
+
+
+def plot_images(
+        images,
+        targets,
+        paths=None,
+        fname='images.jpg',
+        names=None,
+        max_size=640,
+        max_subplots=16):
+    tl = 3  # line thickness
+    tf = max(tl - 1, 1)  # font thickness
+    if os.path.isfile(fname):  # do not overwrite
+        return None
+
+    if isinstance(images, torch.Tensor):
+        images = images.cpu().numpy()
+
+    if isinstance(targets, torch.Tensor):
+        targets = targets.cpu().numpy()
+
+    # un-normalise
+    if np.max(images[0]) <= 1:
+        images *= 255
+
+    bs, _, h, w = images.shape  # batch size, _, height, width
+    bs = min(bs, max_subplots)  # limit plot images
+    ns = np.ceil(bs ** 0.5)  # number of subplots (square)
+
+    # Check if we should resize
+    scale_factor = max_size / max(h, w)
+    if scale_factor < 1:
+        h = math.ceil(scale_factor * h)
+        w = math.ceil(scale_factor * w)
+
+    # Empty array for output
+    mosaic = np.full((int(ns * h), int(ns * w), 3), 255, dtype=np.uint8)
+
+    # Fix class - colour map
+    prop_cycle = plt.rcParams['axes.prop_cycle']
+
+    # https://stackoverflow.com/questions/51350872/python-from-color-name-to-rgb
+    def hex2rgb(h):
+        return tuple(
+            int(h[1 + i:1 + i + 2], 16) for i in (0, 2, 4))
+
+    color_lut = [hex2rgb(h) for h in prop_cycle.by_key()['color']]
+
+    for i, img in enumerate(images):
+        if i == max_subplots:  # if last batch has fewer images than we expect
+            break
+
+        block_x = int(w * (i // ns))
+        block_y = int(h * (i % ns))
+
+        img = img.transpose(1, 2, 0)
+        if scale_factor < 1:
+            img = cv2.resize(img, (w, h))
+
+        mosaic[block_y:block_y + h, block_x:block_x + w, :] = img
+        if len(targets) > 0:
+            image_targets = targets[targets[:, 0] == i]
+            boxes = xywh2xyxy(image_targets[:, 2:6]).T
+            classes = image_targets[:, 1].astype('int')
+            gt = image_targets.shape[1] == 6
+            # ground truth if no conf column
+            # check for confidence presence (gt vs pred)
+            conf = None if gt else image_targets[:, 6]
+
+            boxes[[0, 2]] *= w
+            boxes[[0, 2]] += block_x
+            boxes[[1, 3]] *= h
+            boxes[[1, 3]] += block_y
+            for j, box in enumerate(boxes.T):
+                cls = int(classes[j])
+                color = color_lut[cls % len(color_lut)]
+                cls = names[cls] if names else cls
+                if gt or conf[j] > 0.3:  # 0.3 conf thresh
+                    label = '%s' % cls if gt else '%s %.1f' % (cls, conf[j])
+                    plot_one_box(box, mosaic, label=label,
+                                 color=color, line_thickness=tl)
+
+        # Draw image filename labels
+        if paths is not None:
+            label = os.path.basename(paths[i])[:40]  # trim to 40 char
+            t_size = cv2.getTextSize(
+                label, 0, fontScale=tl / 3, thickness=tf)[0]
+            cv2.putText(mosaic, label, (block_x +
+                                        5, block_y +
+                                        t_size[1] +
+                                        5), 0, tl /
+                        3, [220, 220, 220], thickness=tf, lineType=cv2.LINE_AA)
+
+        # Image border
+        cv2.rectangle(mosaic, (block_x, block_y), (block_x + w,
+                                                   block_y + h), (255, 255, 255), thickness=3)
+
+    if fname is not None:
+        mosaic = cv2.resize(mosaic,
+                            (int(ns * w * 0.5),
+                             int(ns * h * 0.5)),
+                            interpolation=cv2.INTER_AREA)
+        cv2.imwrite(fname, cv2.cvtColor(mosaic, cv2.COLOR_BGR2RGB))
+
+    return mosaic
+
+
+def random_affine(img, targets=(), degrees=10, translate=.1,
+                  scale=.1, shear=10, border=0):
+    # targets = [cls, xyxy]
+
+    height = img.shape[0] + border * 2
+    width = img.shape[1] + border * 2
+
+    # Rotation and Scale
+    R = np.eye(3)
+    a = random.uniform(-degrees, degrees)
+    # a += random.choice([-180, -90, 0, 90])
+    # add 90deg rotations to small rotations
+    s = random.uniform(1 - scale, 1 + scale)
+    # s = 2 ** random.uniform(-scale, scale)
+    R[:2] = cv2.getRotationMatrix2D(angle=a,
+                                    center=(img.shape[1] / 2,
+                                            img.shape[0] / 2),
+                                    scale=s)
+
+    # Translation
+    T = np.eye(3)
+    T[0, 2] = (random.uniform(-translate, translate) *
+               img.shape[0] + border)  # x translation (pixels)
+    T[1, 2] = (random.uniform(-translate, translate) *
+               img.shape[1] + border)  # y translation (pixels)
+
+    # Shear
+    S = np.eye(3)
+    S[0, 1] = math.tan(random.uniform(-shear, shear) *
+                       math.pi / 180)  # x shear (deg)
+    S[1, 0] = math.tan(random.uniform(-shear, shear) *
+                       math.pi / 180)  # y shear (deg)
+
+    # Combined rotation matrix
+    M = S @ T @ R  # ORDER IS IMPORTANT HERE!!
+    if (border != 0) or (M != np.eye(3)).any():  # image changed
+        img = cv2.warpAffine(img, M[:2], dsize=(width, height),
+                             flags=cv2.INTER_LINEAR,
+                             borderValue=(114, 114, 114))
+
+    # Transform label coordinates
+    n = len(targets)
+    if n:
+        # warp points
+        xy = np.ones((n * 4, 3))
+        xy[:, :2] = (targets[:, [1, 2, 3, 4, 1, 4, 3, 2]].
+                     reshape(n * 4, 2))
+        # x1y1, x2y2, x1y2, x2y1
+        xy = (xy @ M.T)[:, :2].reshape(n, 8)
+
+        # create new boxes
+        x = xy[:, [0, 2, 4, 6]]
+        y = xy[:, [1, 3, 5, 7]]
+        xy = np.concatenate((x.min(1), y.min(1), x.max(1),
+                             y.max(1))).reshape(4, n).T
+
+        # reject warped points outside of image
+        xy[:, [0, 2]] = xy[:, [0, 2]].clip(0, width)
+        xy[:, [1, 3]] = xy[:, [1, 3]].clip(0, height)
+        w = xy[:, 2] - xy[:, 0]
+        h = xy[:, 3] - xy[:, 1]
+        area = w * h
+        area0 = ((targets[:, 3] - targets[:, 1]) *
+                 (targets[:, 4] - targets[:, 2]))
+        ar = np.maximum(w / (h + 1e-16), h / (w + 1e-16))
+        # aspect ratio
+        i = (w > 4) & (h > 4) & (area / (area0 * s + 1e-16)
+                                 > 0.2) & (ar < 10)
+
+        targets = targets[i]
+        targets[:, 1:5] = xy[i]
+
+    return img, targets
+
+
+def output_to_target(output, width, height):
+    """
+    Convert a YOLO model output to target format
+    [batch_id, class_id, x, y, w, h, conf]
+    """
+    if isinstance(output, torch.Tensor):
+        output = output.cpu().numpy()
+
+    targets = []
+    for i, o in enumerate(output):
+        if o is not None:
+            for pred in o:
+                box = pred[:4]
+                w = (box[2] - box[0]) / width
+                h = (box[3] - box[1]) / height
+                x = box[0] / width + w / 2
+                y = box[1] / height + h / 2
+                conf = pred[4]
+                cls = int(pred[5])
+
+                targets.append([i, cls, x, y, w, h, conf])
+
+    return np.array(targets)
+
+
+def xyxy2xywh(x):
+    # Convert nx4 boxes from [x1, y1, x2, y2] to [x, y, w, h] where
+    # xy1=top-left, xy2=bottom-right
+    y = torch.zeros_like(x) if isinstance(
+        x, torch.Tensor) else np.zeros_like(x)
+    y[:, 0] = (x[:, 0] + x[:, 2]) / 2  # x center
+    y[:, 1] = (x[:, 1] + x[:, 3]) / 2  # y center
+    y[:, 2] = x[:, 2] - x[:, 0]  # width
+    y[:, 3] = x[:, 3] - x[:, 1]  # height
+    return y
+
+
+def coco80_to_coco91_class():  # converts 80-index (val2014) to 91-index (paper)
+    # https://tech.amikelive.com/node-718/what-object-categories-labels-are-in-coco-dataset/
+    x = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 14, 15, 16, 17, 18, 19, 20,
+         21, 22, 23, 24, 25, 27, 28, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40,
+         41, 42, 43, 44, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58,
+         59, 60, 61, 62, 63, 64, 65, 67, 70, 72, 73, 74, 75, 76, 77, 78, 79,
+         80, 81, 82, 84, 85, 86, 87, 88, 89, 90]
+    return x
+
+
+def check_file(file):
+    # Searches for file if not found locally
+    if os.path.isfile(file):
+        return file
+    else:
+        files = glob.glob('./**/' + file, recursive=True)  # find file
+        assert len(files), 'File Not Found: %s' % file  # assert file was found
+        return files[0]  # return first file if multiple found
+
+
+def load_classes(path):
+    # Loads *.names file at 'path'
+    with open(path, 'r') as f:
+        names = f.read().split('\n')
+    # filter removes empty strings (such as last line)
+    return list(filter(None, names))
+
+
+def load_image(self, index):
+    # loads 1 image from dataset, returns img, original hw, resized hw
+    img = self.imgs[index]
+    if img is None:  # not cached
+        path = self.img_files[index]
+        img = cv2.imread(path)  # BGR
+        assert img is not None, 'Image Not Found ' + path
+        h0, w0 = img.shape[:2]  # orig hw
+        r = self.img_size / max(h0, w0)  # resize image to img_size
+        if r != 1:
+            # always resize down, only resize up if training with augmentation
+            interp = cv2.INTER_AREA if r < 1 and not self.augment \
+                else cv2.INTER_LINEAR
+            img = cv2.resize(img, (int(w0 * r), int(h0 * r)),
+                             interpolation=interp)
+        return img, (h0, w0), img.shape[:2]  # img, hw_original, hw_resized
+    else:
+        # img, hw_original, hw_resized
+        return self.imgs[index], self.img_hw0[index], self.img_hw[index]
+
+
+def load_mosaic(self, index):
+    # loads images in a mosaic
+
+    labels4 = []
+    s = self.img_size
+    xc, yc = [int(random.uniform(s * 0.5, s * 1.5))
+              for _ in range(2)]  # mosaic center x, y
+    indices = [index] + [random.randint(0, len(self.labels) - 1)
+                         for _ in range(3)]  # 3 additional image indices
+    for i, index in enumerate(indices):
+        # Load image
+        img, _, (h, w) = load_image(self, index)
+
+        # place img in img4
+        if i == 0:  # top left
+            img4 = np.full((s * 2, s * 2, img.shape[2]),
+                           114, dtype=np.uint8)
+            # base image with 4 tiles
+            x1a, y1a, x2a, y2a = (max(xc - w, 0),
+                                  max(yc - h, 0), xc, yc)
+            # xmin, ymin, xmax, ymax (large image)
+            x1b, y1b, x2b, y2b = (w - (x2a - x1a), h -
+                                  (y2a - y1a), w, h)
+            # xmin, ymin, xmax, ymax (small image)
+        elif i == 1:  # top right
+            x1a, y1a, x2a, y2a = (xc, max(yc - h, 0),
+                                  min(xc + w, s * 2), yc)
+            x1b, y1b, x2b, y2b = (0, h - (y2a - y1a),
+                                  min(w, x2a - x1a), h)
+        elif i == 2:  # bottom left
+            x1a, y1a, x2a, y2a = (max(xc - w, 0), yc,
+                                  xc, min(s * 2, yc + h))
+            x1b, y1b, x2b, y2b = (w - (x2a - x1a), 0,
+                                  max(xc, w), min(y2a - y1a, h))
+        elif i == 3:  # bottom right
+            x1a, y1a, x2a, y2a = xc, yc, min(xc + w,
+                                             s * 2), min(s * 2, yc + h)
+            x1b, y1b, x2b, y2b = (0, 0,
+                                  min(w, x2a - x1a), min(y2a - y1a, h))
+
+        img4[y1a:y2a, x1a:x2a] = img[y1b:y2b, x1b:x2b]
+        # img4[ymin:ymax, xmin:xmax]
+        padw = x1a - x1b
+        padh = y1a - y1b
+
+        # Labels
+        x = self.labels[index]
+        labels = x.copy()
+        if x.size > 0:  # Normalized xywh to pixel xyxy format
+            labels[:, 1] = w * (x[:, 1] - x[:, 3] / 2) + padw
+            labels[:, 2] = h * (x[:, 2] - x[:, 4] / 2) + padh
+            labels[:, 3] = w * (x[:, 1] + x[:, 3] / 2) + padw
+            labels[:, 4] = h * (x[:, 2] + x[:, 4] / 2) + padh
+        labels4.append(labels)
+
+    # Concat/clip labels
+    if len(labels4):
+        labels4 = np.concatenate(labels4, 0)
+        # np.clip(labels4[:, 1:] - s / 2, 0, s, out=labels4[:, 1:])
+        # use with center crop
+        np.clip(labels4[:, 1:], 0, 2 * s, out=labels4[:, 1:])
+        # use with random_affine
+
+    # Augment
+    # img4 = img4[s // 2: int(s * 1.5), s // 2:int(s * 1.5)]
+    # center crop (WARNING, requires box pruning)
+    img4, labels4 = random_affine(img4, labels4,
+                                  degrees=self.hyp['degrees'],
+                                  translate=self.hyp['translate'],
+                                  scale=self.hyp['scale'],
+                                  shear=self.hyp['shear'],
+                                  border=-s // 2)  # border to remove
+
+    return img4, labels4
+
+
+def compute_ap(recall, precision):
+    """ Compute the average precision, given the recall and precision curves.
+    Source: https://github.com/rbgirshick/py-faster-rcnn.
+    # Arguments
+        recall:    The recall curve (list).
+        precision: The precision curve (list).
+    # Returns
+        The average precision as computed in py-faster-rcnn.
+    """
+
+    # Append sentinel values to beginning and end
+    mrec = np.concatenate(([0.], recall, [min(recall[-1] + 1E-3, 1.)]))
+    mpre = np.concatenate(([0.], precision, [0.]))
+
+    # Compute the precision envelope
+    mpre = np.flip(np.maximum.accumulate(np.flip(mpre)))
+
+    # Integrate area under curve
+    method = 'interp'  # methods: 'continuous', 'interp'
+    if method == 'interp':
+        x = np.linspace(0, 1, 101)  # 101-point interp (COCO)
+        ap = np.trapz(np.interp(x, mrec, mpre), x)  # integrate
+    else:  # 'continuous'
+        # points where x axis (recall) changes
+        i = np.where(mrec[1:] != mrec[:-1])[0]
+        ap = np.sum((mrec[i + 1] - mrec[i]) * mpre[i + 1])  # area under curve
+
+    return ap
+
+
+def augment_hsv(img, hgain=0.5, sgain=0.5, vgain=0.5):
+    r = (np.random.uniform(-1, 1, 3) *
+         [hgain, sgain, vgain] + 1)  # random gains
+    hue, sat, val = cv2.split(cv2.cvtColor(img, cv2.COLOR_BGR2HSV))
+    dtype = img.dtype  # uint8
+
+    x = np.arange(0, 256, dtype=np.int16)
+    lut_hue = ((x * r[0]) % 180).astype(dtype)
+    lut_sat = np.clip(x * r[1], 0, 255).astype(dtype)
+    lut_val = np.clip(x * r[2], 0, 255).astype(dtype)
+
+    img_hsv = cv2.merge((cv2.LUT(hue, lut_hue),
+                         cv2.LUT(sat, lut_sat),
+                         cv2.LUT(val, lut_val))).astype(dtype)
+    cv2.cvtColor(img_hsv, cv2.COLOR_HSV2BGR, dst=img)
+    # no return needed
+
+
+class LoadImagesAndLabels(Dataset):  # for training/testing
+    def __init__(
+            self,
+            path,
+            img_size=416,
+            batch_size=16,
+            augment=False,
+            hyp=None,
+            rect=False,
+            image_weights=False,
+            cache_images=False,
+            single_cls=False,
+            pad=0.0):
+        try:
+            path = str(Path(path))  # os-agnostic
+            parent = str(Path(path).parent) + os.sep
+            if os.path.isfile(path):  # file
+                with open(path, 'r') as f:
+                    f = f.read().splitlines()
+                    # local to global path
+                    f = [
+                        x.replace(
+                            './',
+                            parent) if x.startswith('./') else x for x in f]
+            elif os.path.isdir(path):  # folder
+                f = glob.iglob(path + os.sep + '*.*')
+            else:
+                raise Exception('%s does not exist' % path)
+            self.img_files = [x.replace(
+                '/', os.sep) for x in f if
+                os.path.splitext(x)[-1].lower() in img_formats]
+
+        except BaseException:
+            raise Exception(
+                'Error loading data from %s. See %s' %
+                (path, help_url))
+
+        n = len(self.img_files)
+        assert n > 0, 'No images found in %s. See %s' % (path, help_url)
+        bi = np.floor(np.arange(n) / batch_size).astype(int)  # batch index
+        nb = bi[-1] + 1  # number of batches
+
+        self.n = n  # number of images
+        self.batch = bi  # batch index of image
+        self.img_size = img_size
+        self.augment = augment
+        self.hyp = hyp
+        self.image_weights = image_weights
+        self.rect = False if image_weights else rect
+        # load 4 images at a time into a mosaic (only during training)
+        self.mosaic = self.augment and not self.rect
+
+        # Define labels
+        self.label_files = [x.replace('images', 'labels').replace(
+            os.path.splitext(x)[-1], '.txt') for x in self.img_files]
+
+        # Read image shapes (wh)
+        sp = path.replace('.txt', '') + '.shapes'  # shapefile path
+        try:
+            with open(sp, 'r') as f:  # read existing shapefile
+                s = [x.split() for x in f.read().splitlines()]
+                assert len(s) == n, 'Shapefile out of sync'
+        except BaseException:
+            s = [exif_size(Image.open(f)) for f in tqdm(
+                self.img_files,
+                desc='Reading image shapes')]
+            np.savetxt(sp, s, fmt='%g')  # overwrites existing (if any)
+
+        self.shapes = np.array(s, dtype=np.float64)
+
+        # Rectangular Training
+        # https://github.com/ultralytics/yolov3/issues/232
+        if self.rect:
+            # Sort by aspect ratio
+            s = self.shapes  # wh
+            ar = s[:, 1] / s[:, 0]  # aspect ratio
+            irect = ar.argsort()
+            self.img_files = [self.img_files[i] for i in irect]
+            self.label_files = [self.label_files[i] for i in irect]
+            self.shapes = s[irect]  # wh
+            ar = ar[irect]
+
+            # Set training image shapes
+            shapes = [[1, 1]] * nb
+            for i in range(nb):
+                ari = ar[bi == i]
+                mini, maxi = ari.min(), ari.max()
+                if maxi < 1:
+                    shapes[i] = [maxi, 1]
+                elif mini > 1:
+                    shapes[i] = [1, 1 / mini]
+
+            self.batch_shapes = np.ceil(
+                np.array(shapes) * img_size / 32. + pad).astype(int) * 32
+
+        # Cache labels
+        self.imgs = [None] * n
+        self.labels = [np.zeros((0, 5), dtype=np.float32)] * n
+        create_datasubset, extract_bounding_boxes, labels_loaded = \
+            False, False, False
+        # number missing, found, empty, datasubset, duplicate
+        nm, nf, ne, ns, nd = 0, 0, 0, 0, 0
+        # saved labels in *.npy file
+        np_labels_path = str(Path(self.label_files[0]).parent) + '.npy'
+        if os.path.isfile(np_labels_path):
+            s = np_labels_path  # print string
+
+            print(np_labels_path)
+
+            x = np.load(np_labels_path, allow_pickle=True)
+            if len(x) == n:
+                self.labels = x
+                labels_loaded = True
+        else:
+            s = path.replace('images', 'labels')
+
+        pbar = tqdm(self.label_files)
+        for i, file in enumerate(pbar):
+            if labels_loaded:
+                l = self.labels[i]
+                # np.savetxt(file, l, '%g')  # save *.txt from *.npy file
+            else:
+                try:
+                    with open(file, 'r') as f:
+                        l = np.array(
+                            [x.split() for x in f.read().splitlines()],
+                            dtype=np.float32)
+                except BaseException:
+                    # print('missing labels for image %s' % self.img_files[i])
+                    # # file missing
+                    nm += 1
+                    continue
+
+            if l.shape[0]:
+                assert l.shape[1] == 5, '> 5 label columns: %s' % file
+                assert (l >= 0).all(), 'negative labels: %s' % file
+                assert (l[:, 1:] <= 1).all(
+                ), 'non-normalized or out of bounds coordinate labels: %s' % file
+                if np.unique(
+                        l, axis=0).shape[0] < l.shape[0]:  # duplicate rows
+                    # print('WARNING: duplicate rows in %s' %
+                    # self.label_files[i])  # duplicate rows
+                    nd += 1
+                if single_cls:
+                    l[:, 0] = 0  # force dataset into single-class mode
+                self.labels[i] = l
+                nf += 1  # file found
+
+                # Create subdataset (a smaller dataset)
+                if create_datasubset and ns < 1E4:
+                    if ns == 0:
+                        create_folder(path='./datasubset')
+                        os.makedirs('./datasubset/images')
+                    exclude_classes = 43
+                    if exclude_classes not in l[:, 0]:
+                        ns += 1
+                        # shutil.copy(src=self.img_files[i],
+                        # dst='./datasubset/images/')  # copy image
+                        with open('./datasubset/images.txt', 'a') as f:
+                            f.write(self.img_files[i] + '\n')
+
+                # Extract object detection boxes for a second stage classifier
+                if extract_bounding_boxes:
+                    p = Path(self.img_files[i])
+                    img = cv2.imread(str(p))
+                    h, w = img.shape[:2]
+                    for j, x in enumerate(l):
+                        f = '%s%sclassifier%s%g_%g_%s' % (
+                            p.parent.parent, os.sep, os.sep, x[0], j, p.name)
+                        if not os.path.exists(Path(f).parent):
+                            # make new output folder
+                            os.makedirs(Path(f).parent)
+
+                        b = x[1:] * [w, h, w, h]  # box
+                        b[2:] = b[2:].max()  # rectangle to square
+                        b[2:] = b[2:] * 1.3 + 30  # pad
+                        b = xywh2xyxy(b.reshape(-1, 4)).ravel().astype(int)
+
+                        # clip boxes outside of image
+                        b[[0, 2]] = np.clip(b[[0, 2]], 0, w)
+                        b[[1, 3]] = np.clip(b[[1, 3]], 0, h)
+                        assert cv2.imwrite(
+                            f, img[b[1]:b[3], b[0]:b[2]]), \
+                            'Failure extracting classifier boxes'
+            else:
+                # print('empty labels for image %s' % self.img_files[i])  #
+                # file empty
+                ne += 1
+                # os.system("rm '%s' '%s'" % (self.img_files[i],
+                # self.label_files[i]))  # remove
+
+            pbar.desc = 'Caching labels %s (%g found, %g missing, %g empty,\
+             %g duplicate, for %g images)' % (
+                s, nf, nm, ne, nd, n)
+        assert nf > 0 or n == 20288, 'No labels found in %s. See %s' % (
+            os.path.dirname(file) + os.sep, help_url)
+        if not labels_loaded and n > 1000:
+            print(
+                'Saving labels to %s for faster future loading' %
+                np_labels_path)
+            # np.save(np_labels_path, self.labels)  # save for next time
+
+        # Cache images into memory for faster training (WARNING: large datasets
+        # may exceed system RAM)
+        if cache_images:  # if training
+            gb = 0  # Gigabytes of cached images
+            pbar = tqdm(range(len(self.img_files)), desc='Caching images')
+            self.img_hw0, self.img_hw = [None] * n, [None] * n
+            for i in pbar:  # max 10k images
+                self.imgs[i], self.img_hw0[i], self.img_hw[i] = load_image(
+                    self, i)  # img, hw_original, hw_resized
+                gb += self.imgs[i].nbytes
+                pbar.desc = 'Caching images (%.1fGB)' % (gb / 1E9)
+
+    def __len__(self):
+        return len(self.img_files)
+
+    def __getitem__(self, index):
+        if self.image_weights:
+            index = self.indices[index]
+
+        hyp = self.hyp
+        if self.mosaic:
+            # Load mosaic
+            img, labels = load_mosaic(self, index)
+            shapes = None
+
+        else:
+            # Load image
+            img, (h0, w0), (h, w) = load_image(self, index)
+
+            # Letterbox
+            shape = self.batch_shapes[self.batch[index]] if self.rect else self.img_size  # final letterboxed shape
+            img, ratio, pad = letterbox(img, shape, auto=False, scaleup=self.augment)
+            shapes = (h0, w0), ((h / h0, w / w0), pad)  # for COCO mAP rescaling
+
+            # Load labels
+            labels = []
+            x = self.labels[index]
+            if x.size > 0:
+                # Normalized xywh to pixel xyxy format
+                labels = x.copy()
+                labels[:, 1] = ratio[0] * w * (x[:, 1] - x[:, 3] / 2) + pad[0]  # pad width
+                labels[:, 2] = ratio[1] * h * (x[:, 2] - x[:, 4] / 2) + pad[1]  # pad height
+                labels[:, 3] = ratio[0] * w * (x[:, 1] + x[:, 3] / 2) + pad[0]
+                labels[:, 4] = ratio[1] * h * (x[:, 2] + x[:, 4] / 2) + pad[1]
+
+        if self.augment:
+            # Augment imagespace
+            if not self.mosaic:
+                img, labels = random_affine(img, labels,
+                                            degrees=hyp['degrees'],
+                                            translate=hyp['translate'],
+                                            scale=hyp['scale'],
+                                            shear=hyp['shear'])
+
+            # Augment colorspace
+            augment_hsv(img, hgain=hyp['hsv_h'], sgain=hyp['hsv_s'], vgain=hyp['hsv_v'])
+
+            # Apply cutouts
+            # if random.random() < 0.9:
+            #     labels = cutout(img, labels)
+
+        nL = len(labels)  # number of labels
+        if nL:
+            # convert xyxy to xywh
+            labels[:, 1:5] = xyxy2xywh(labels[:, 1:5])
+
+            # Normalize coordinates 0 - 1
+            labels[:, [2, 4]] /= img.shape[0]  # height
+            labels[:, [1, 3]] /= img.shape[1]  # width
+
+        if self.augment:
+            # random left-right flip
+            lr_flip = True
+            if lr_flip and random.random() < 0.5:
+                img = np.fliplr(img)
+                if nL:
+                    labels[:, 1] = 1 - labels[:, 1]
+
+            # random up-down flip
+            ud_flip = False
+            if ud_flip and random.random() < 0.5:
+                img = np.flipud(img)
+                if nL:
+                    labels[:, 2] = 1 - labels[:, 2]
+
+        labels_out = torch.zeros((nL, 6))
+        if nL:
+            labels_out[:, 1:] = torch.from_numpy(labels)
+
+        # Convert
+        img = img[:, :, ::-1].transpose(2, 0, 1)  # BGR to RGB, to 3x416x416
+        img = np.ascontiguousarray(img)
+
+        return torch.from_numpy(img), labels_out, self.img_files[index], shapes
+
+    @staticmethod
+    def collate_fn(batch):
+        img, label, path, shapes = zip(*batch)  # transposed
+        for i, l in enumerate(label):
+            l[:, 0] = i  # add target image index for build_targets()
+        return torch.stack(img, 0), torch.cat(label, 0), path, shapes
+
+
+def parse_data_cfg(path):
+    # Parses the data configuration file
+    if not os.path.exists(path) and os.path.exists(
+            'data' + os.sep + path):  # add data/ prefix if omitted
+        path = 'data' + os.sep + path
+
+    with open(path, 'r') as f:
+        lines = f.readlines()
+
+    options = dict()
+    for line in lines:
+        line = line.strip()
+        if line == '' or line.startswith('#'):
+            continue
+        key, val = line.split('=')
+        options[key.strip()] = val.strip()
+
+    return options
+
+
+def create_grids(ng=(13, 13), device='cpu'):
+    nx, ny = ng  # x and y grid size
+    ng = torch.tensor(ng, dtype=torch.float)
+
+    # build xy offsets
+    yv, xv = torch.meshgrid(
+        [torch.arange(ny, device=device), torch.arange(nx, device=device)])
+    grid = torch.stack((xv, yv), 2).view((1, 1, ny, nx, 2)).float()
+
+    return grid
+
+
+def post_process(x):
+    stride = [32, 16, 8]
+    anchors = [[10, 13, 16, 30, 33, 23],
+               [30, 61, 62, 45, 59, 119],
+               [116, 90, 156, 198, 373, 326]]
+    temp = [13, 26, 52]
+
+    res = []
+    for i in range(3):
+        out = torch.from_numpy(x[i]) if not torch.is_tensor(x[i]) else x[i]
+
+        bs, _, ny, nx = out.shape  # bs, 255, 13, 13
+
+        anchor = torch.Tensor(anchors[2 - i]).reshape(3, 2)
+        anchor_vec = anchor / stride[i]
+        anchor_wh = anchor_vec.view(1, 3, 1, 1, 2)
+
+        grid = create_grids((nx, ny))
+
+        # p.view(bs, 255, 13, 13) -- > (bs, 3, 13, 13, 85)  # (bs, anchors,
+        # grid, grid, classes + xywh)
+        out = out.view(
+            bs, 3, 85, temp[i], temp[i]).permute(
+            0, 1, 3, 4, 2).contiguous()  # prediction
+
+        io = out.clone()  # inference output
+
+        io[..., :2] = torch.sigmoid(io[..., :2]) + grid  # xy
+        io[..., 2:4] = torch.exp(io[..., 2:4]) * anchor_wh  # wh yolo method
+        io[..., :4] *= stride[i]
+        torch.sigmoid_(io[..., 4:])
+
+        res.append(io.view(bs, -1, 85))
+    return torch.cat(res, 1), x
+
+
+def test(data,
+         batch_size=32,
+         imgsz=416,
+         conf_thres=0.001,
+         iou_thres=0.6,  # for nms
+         save_json=False,
+         single_cls=False,
+         augment=False,
+         model=None,
+         dataloader=None,
+         multi_label=True,
+         names='data/coco.names',
+         onnx_runtime=True,
+         onnx_weights="yolov3-8",
+         ipu=False,
+         provider_config='vaip_config.json'):
+    """
+    COCO average precision (AP) Evaluation. Iterate inference on the test dataset
+    and the results are evaluated by COCO API.
+    """
+
+    device = torch.device('cpu')
+    verbose = False
+    if isinstance(onnx_weights, list):
+        onnx_weights = onnx_weights[0]
+
+    if ipu:
+        providers = ["VitisAIExecutionProvider"]
+        provider_options = [{"config_file": provider_config}]
+    else:
+        providers = ['CUDAExecutionProvider', 'CPUExecutionProvider']
+        provider_options = None
+
+    onnx_model = onnxruntime.InferenceSession(
+        onnx_weights,
+        providers=providers,
+        provider_options=provider_options)
+
+    # Configure run
+    data = parse_data_cfg(data)
+    nc = 1 if single_cls else int(data['classes'])  # number of classes
+    path = data['valid']  # path to test images
+    names = load_classes(data['names'])  # class names
+    iouv = torch.linspace(0.5, 0.95, 10).to(
+        device)  # iou vector for mAP@0.5:0.95
+    iouv = iouv[0].view(1)  # comment for mAP@0.5:0.95
+    niou = iouv.numel()
+
+    # Dataloader
+    if dataloader is None:
+        dataset = LoadImagesAndLabels(
+            path,
+            imgsz,
+            batch_size,
+            rect=False,
+            single_cls=opt.single_cls,
+            pad=0.5)
+        batch_size = min(batch_size, len(dataset))
+        dataloader = DataLoader(dataset,
+                                batch_size=batch_size,
+                                num_workers=min([os.cpu_count(),
+                                                 batch_size if
+                                                 batch_size > 1 else 0,
+                                                 8]),
+                                pin_memory=True,
+                                collate_fn=dataset.collate_fn)
+
+    seen = 0
+
+    coco91class = coco80_to_coco91_class()
+    s = ('%20s' + '%10s' * 6) % ('Class', 'Images', 'Targets', 'P', 'R',
+                                 'mAP@0.5', 'F1')
+    p, r, f1, mp, mr, map, mf1, t0, t1 = 0., 0., 0., 0., 0., 0., 0., 0., 0.
+    loss = torch.zeros(3, device=device)
+    jdict, stats, ap, ap_class = [], [], [], []
+
+    for batch_i, (imgs, targets, paths, shapes) in enumerate(
+            tqdm(dataloader, desc=s)):
+        # uint8 to float32, 0 - 255 to 0.0 - 1.0
+        imgs = imgs.to(device).float() / 255.0
+        targets = targets.to(device)
+        nb, _, height, width = imgs.shape
+        # batch size, channels, height, width
+        whwh = torch.Tensor([width, height, width, height]).to(device)
+
+        if onnx_runtime:
+            outputs = onnx_model.run(
+                None, {onnx_model.get_inputs()[0].name: imgs.cpu().numpy()})
+            outputs = [torch.tensor(item).to(device) for item in outputs]
+            inf_out, train_out = post_process(outputs)
+
+        else:
+
+            # Disable gradients
+            with torch.no_grad():
+                # Run model
+                t = time_synchronized()
+
+                # inference and training outputs
+                inf_out, train_out = model(imgs, augment=augment)
+                t0 += time_synchronized() - t
+
+        # Compute loss
+        # if is_training:  # if model has loss hyperparameters
+        # loss += compute_loss(train_out, targets, model)[1][:3]  # GIoU,
+        # obj, cls
+
+        # Run NMS
+        t = time_synchronized()
+        output = non_max_suppression(
+            inf_out,
+            conf_thres=conf_thres,
+            iou_thres=iou_thres,
+            multi_label=multi_label)
+        t1 += time_synchronized() - t
+
+        # Statistics per image
+        for si, pred in enumerate(output):
+            labels = targets[targets[:, 0] == si, 1:]
+            nl = len(labels)
+            tcls = labels[:, 0].tolist() if nl else []  # target class
+            seen += 1
+
+            if pred is None:
+                if nl:
+                    stats.append(
+                        (torch.zeros(
+                            0,
+                            niou,
+                            dtype=torch.bool),
+                         torch.Tensor(),
+                         torch.Tensor(),
+                         tcls))
+                continue
+
+            # Append to text file
+            # with open('test.txt', 'a') as file:
+            #    [file.write('%11.5g' * 7 % tuple(x) + '\n') for x in pred]
+
+            # Clip boxes to image bounds
+            clip_coords(pred, (height, width))
+
+            # Append to pycocotools JSON dictionary
+            if save_json:
+                image_id = int(Path(paths[si]).stem.split('_')[-1])
+                box = pred[:, :4].clone()  # xyxy
+                scale_coords(imgs[si].shape[1:], box, shapes[si]
+                [0], shapes[si][1])  # to original shape
+                box = xyxy2xywh(box)  # xywh
+                box[:, :2] -= box[:, 2:] / 2  # xy center to top-left corner
+                for p, b in zip(pred.tolist(), box.tolist()):
+                    jdict.append({'image_id': image_id,
+                                  'category_id': coco91class[int(p[5])],
+                                  'bbox': [round(x, 3) for x in b],
+                                  'score': round(p[4], 5)})
+
+            # Assign all predictions as incorrect
+            correct = torch.zeros(
+                pred.shape[0],
+                niou,
+                dtype=torch.bool,
+                device=device)
+            if nl:
+                detected = []  # target indices
+                tcls_tensor = labels[:, 0]
+
+                # target boxes
+                tbox = xywh2xyxy(labels[:, 1:5]) * whwh
+
+                # Per target class
+                for cls in torch.unique(tcls_tensor):
+                    ti = (cls == tcls_tensor).nonzero(
+                    ).view(-1)  # target indices
+                    pi = (cls == pred[:, 5]).nonzero(
+                    ).view(-1)  # prediction indices
+
+                    # Search for detections
+                    if pi.shape[0]:
+                        # Prediction to target ious
+                        ious, i = box_iou(pred[pi, :4], tbox[ti].cpu()).max(
+                            1)  # best ious, indices
+
+                        # Append detections
+                        for j in (ious > iouv[0].cpu()).nonzero():
+                            d = ti[i[j]]  # detected target
+                            if d not in detected:
+                                detected.append(d)
+                                # iou_thres is 1xn
+                                correct[pi[j]] = ious[j] > iouv.cpu()
+                                if len(
+                                        detected) == nl:
+                                    # all targets already located in image
+                                    break
+
+            # Append statistics (correct, conf, pcls, tcls)
+            stats.append(
+                (correct.cpu(), pred[:, 4].cpu(), pred[:, 5].cpu(), tcls))
+
+        # Plot images
+        if batch_i < 1:
+            f = 'test_batch%g_gt.jpg' % batch_i  # filename
+            plot_images(imgs, targets, paths=paths, names=names,
+                        fname=f)  # ground truth
+            f = 'test_batch%g_pred.jpg' % batch_i
+            plot_images(imgs, output_to_target(output, width, height),
+                        paths=paths, names=names, fname=f)  # predictions
+
+    # test end
+
+    # Compute statistics
+    stats = [np.concatenate(x, 0) for x in zip(*stats)]  # to numpy
+    if len(stats):
+        p, r, ap, f1, ap_class = ap_per_class(*stats)
+        if niou > 1:
+            p, r, ap, f1 = p[:, 0], r[:, 0], ap.mean(
+                1), ap[:, 0]  # [P, R, AP@0.5:0.95, AP@0.5]
+        mp, mr, map, mf1 = p.mean(), r.mean(), ap.mean(), f1.mean()
+        nt = np.bincount(stats[3].astype(np.int64),
+                         minlength=nc)  # number of targets per class
+    else:
+        nt = torch.zeros(1)
+
+    # Print results
+    pf = '%20s' + '%10.3g' * 6  # print format
+    print(pf % ('all', seen, nt.sum(), mp, mr, map, mf1))
+
+    # Print results per class
+    if verbose and nc > 1 and len(stats):
+        for i, c in enumerate(ap_class):
+            print(pf % (names[c], seen, nt[c], p[i], r[i], ap[i], f1[i]))
+
+    # Print speeds
+    if verbose or save_json:
+        t = tuple(x / seen * 1E3 for x in (t0, t1, t0 + t1)) + \
+            (imgsz, imgsz, batch_size)  # tuple
+        print(
+            'Speed: %.1f/%.1f/%.1f ms \
+            inference/NMS/total per %gx%g image at batch-size %g' % t)
+
+    # Save JSON
+    if save_json and map and len(jdict):
+        print('\nCOCO mAP with pycocotools...')
+        imgIds = [int(Path(x).stem.split('_')[-1])
+                  for x in dataloader.dataset.img_files]
+        with open('results.json', 'w') as file:
+            json.dump(jdict, file)
+
+        try:
+            from pycocotools.coco import COCO
+            from pycocotools.cocoeval import COCOeval
+
+            # https://github.com/cocodataset/cocoapi/blob/master/PythonAPI/pycocoEvalDemo.ipynb
+            # initialize COCO ground truth api
+            cocoGt = COCO(
+                glob.glob('coco/annotations/instances_val*.json')[0])
+            cocoDt = cocoGt.loadRes('results.json')  # initialize COCO pred api
+            cocoEval = COCOeval(cocoGt, cocoDt, 'bbox')
+            # [:32]  # only evaluate these images
+            cocoEval.params.imgIds = imgIds
+            cocoEval.evaluate()
+            cocoEval.accumulate()
+            cocoEval.summarize()
+            # mf1, map = cocoEval.stats[:2]  # update to pycocotools results
+            # (mAP@0.5:0.95, mAP@0.5)
+        except BaseException:
+            print(
+                'WARNING: pycocotools must be installed with \
+                numpy==1.17 to run correctly. '
+                'See https://github.com/cocodataset/cocoapi/issues/356')
+
+    # Return results
+    maps = np.zeros(nc) + map
+    for i, c in enumerate(ap_class):
+        maps[c] = ap[i]
+    return (mp, mr, map, mf1, *(loss.cpu() / len(dataloader)).tolist()), maps
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(prog='Test onnx model performance on COCO dataset')
+    parser.add_argument(
+        '--data',
+        type=str,
+        default='coco2017.data',
+        help='Path of *.data')
+    parser.add_argument(
+        '--batch-size',
+        type=int,
+        default=1,
+        help='Size of each image batch')
+    parser.add_argument(
+        '--img-size',
+        type=int,
+        default=416,
+        help='Inference size (pixels)')
+    parser.add_argument(
+        '--conf-thres',
+        type=float,
+        default=0.001,
+        help='Object confidence threshold')
+    parser.add_argument(
+        '--iou-thres',
+        type=float,
+        default=0.5,
+        help='IOU threshold for NMS')
+    parser.add_argument(
+        '--save-json',
+        action='store_true',
+        help='Save a COCOapi-compatible JSON results file')
+    parser.add_argument(
+        '--device',
+        default='',
+        help='Device id (i.e. 0 or 0,1) or cpu')
+    parser.add_argument(
+        '--augment',
+        action='store_true',
+        help='Augmented inference')
+    parser.add_argument('--sync_bn', action='store_true')
+    parser.add_argument('--print_model', action='store_true')
+    parser.add_argument('--test_rect', action='store_true')
+
+    parser.add_argument(
+        '--onnx_runtime',
+        action='store_true',
+        help='Use onnx runtime')
+    parser.add_argument(
+        '--onnx_weights',
+        default='yolov3-8.onnx',
+        nargs='+',
+        type=str,
+        help='Path of onnx weights')
+    parser.add_argument(
+        '--single-cls',
+        action='store_true',
+        help='Run as single-class dataset')
+    parser.add_argument(
+        "--ipu",
+        action="store_true",
+        help="Use IPU for inference")
+    parser.add_argument(
+        "--provider_config",
+        type=str,
+        default="vaip_config.json",
+        help="Path of the config file for seting provider_options")
+
+    opt = parser.parse_args()
+    opt.save_json = opt.save_json or any(
+        [x in opt.data for x in ['coco.data',
+                                 'coco2014.data', 'coco2017.data']])
+    opt.data = check_file(opt.data)  # check file
+    print(opt)
+
+    help_url = 'https://github.com/ultralytics/yolov3/wiki/Train-Custom-Data'
+    img_formats = ['.bmp', '.jpg', '.jpeg', '.png', '.tif', '.tiff', '.dng']
+    vid_formats = ['.mov', '.avi', '.mp4', '.mpg', '.mpeg', '.m4v', '.wmv',
+                   '.mkv']
+
+    test(opt.data,
+         opt.batch_size,
+         opt.img_size,
+         opt.conf_thres,
+         opt.iou_thres,
+         opt.save_json,
+         opt.single_cls,
+         opt.augment,
+         names='data/coco.names',
+         onnx_weights=opt.onnx_weights,
+         ipu=opt.ipu,
+         provider_config=opt.provider_config
+         )

requirements.txt

@@ -0,0 +1,29 @@
+# pip install -r requirements.txt
+
+# base ----------------------------------------
+Cython
+matplotlib>=3.2.2
+numpy>=1.18.5
+opencv-python>=4.1.2
+pillow
+PyYAML>=5.3
+scipy>=1.4.1
+tensorboard>=2.2
+torch==1.12.0
+torchvision>=0.7.0
+tqdm>=4.41.0
+pandas
+#onnxruntime
+
+# coco ----------------------------------------
+pycocotools>=2.0
+
+# export --------------------------------------
+# packaging  # for coremltools
+# coremltools==4.0b3
+# onnx>=1.7.0
+# scikit-learn==0.19.2  # for coreml quantization
+
+# extras --------------------------------------
+# thop  # FLOPS computation
+# seaborn  # plotting

utils.py

@@ -0,0 +1,213 @@
+import numpy as np
+import cv2
+import torch
+import time
+import torchvision
+import random
+
+
+def box_iou(box1, box2):
+    # https://github.com/pytorch/vision/blob/master/torchvision/ops/boxes.py
+    """
+    Return intersection-over-union (Jaccard index) of boxes.
+    Both sets of boxes are expected to be in (x1, y1, x2, y2) format.
+    Arguments:
+        box1 (Tensor[N, 4])
+        box2 (Tensor[M, 4])
+    Returns:
+        iou (Tensor[N, M]): the NxM matrix containing the pairwise
+            IoU values for every element in boxes1 and boxes2
+    """
+
+    def box_area(box):
+        # box = 4xn
+        return (box[2] - box[0]) * (box[3] - box[1])
+
+    area1 = box_area(box1.T)
+    area2 = box_area(box2.T)
+
+    # inter(N,M) = (rb(N,M,2) - lt(N,M,2)).clamp(0).prod(2)
+    inter = (torch.min(box1[:, None, 2:], box2[:, 2:]) - torch.max(box1[:, None, :2], box2[:, :2])).clamp(0).prod(2)
+    return inter / (area1[:, None] + area2 - inter)  # iou = inter / (area1 + area2 - inter)
+
+
+def plot_one_box(x, image, color=None, label=None, line_thickness=None):
+    # Plots one bounding box on image img
+    tl = line_thickness or round(
+        0.002 * (image.shape[0] + image.shape[1]) / 2) + 1  # line/font thickness
+    color = color or [random.randint(0, 255) for _ in range(3)]
+    c1, c2 = (int(x[0]), int(x[1])), (int(x[2]), int(x[3]))
+    cv2.rectangle(image, c1, c2, color, thickness=tl, lineType=cv2.LINE_AA)
+    if label:
+        tf = max(tl - 1, 1)  # font thickness
+        t_size = cv2.getTextSize(label, 0, fontScale=tl / 3, thickness=tf)[0]
+        c2 = c1[0] + t_size[0], c1[1] - t_size[1] - 3
+        cv2.rectangle(image, c1, c2, color, -1, cv2.LINE_AA)  # filled
+        cv2.putText(image, label, (c1[0], c1[1] - 2), 0, tl / 3,
+                    [225, 255, 255], thickness=tf, lineType=cv2.LINE_AA)
+
+
+def clip_coords(boxes, img_shape):
+    # Clip bounding xyxy bounding boxes to image shape (height, width)
+    boxes[:, 0].clamp_(0, img_shape[1])  # x1
+    boxes[:, 1].clamp_(0, img_shape[0])  # y1
+    boxes[:, 2].clamp_(0, img_shape[1])  # x2
+    boxes[:, 3].clamp_(0, img_shape[0])  # y2
+
+
+def scale_coords(img1_shape, coords, img0_shape, ratio_pad=None):
+    # Rescale coords (xyxy) from img1_shape to img0_shape
+    if ratio_pad is None:  # calculate from img0_shape
+        gain = max(img1_shape) / max(img0_shape)  # gain  = old / new
+        pad = (img1_shape[1] - img0_shape[1] * gain) / \
+              2, (img1_shape[0] - img0_shape[0] * gain) / 2  # wh padding
+    else:
+        gain = ratio_pad[0][0]
+        pad = ratio_pad[1]
+
+    coords[:, [0, 2]] -= pad[0]  # x padding
+    coords[:, [1, 3]] -= pad[1]  # y padding
+    coords[:, :4] /= gain
+    clip_coords(coords, img0_shape)
+    return coords
+
+
+def xywh2xyxy(x):
+    # Convert nx4 boxes from [x, y, w, h] to [x1, y1, x2, y2] where
+    # xy1=top-left, xy2=bottom-right
+    y = torch.zeros_like(x) if isinstance(
+        x, torch.Tensor) else np.zeros_like(x)
+    y[:, 0] = x[:, 0] - x[:, 2] / 2  # top left x
+    y[:, 1] = x[:, 1] - x[:, 3] / 2  # top left y
+    y[:, 2] = x[:, 0] + x[:, 2] / 2  # bottom right x
+    y[:, 3] = x[:, 1] + x[:, 3] / 2  # bottom right y
+    return y
+
+
+def letterbox(img, new_shape=(416, 416), color=(114, 114, 114), auto=True,
+              scaleFill=False, scaleup=True):
+    # Resize image to a 32-pixel-multiple rectangle
+    # https://github.com/ultralytics/yolov3/issues/232
+    shape = img.shape[:2]  # current shape [height, width]
+    if isinstance(new_shape, int):
+        new_shape = (new_shape, new_shape)
+
+    # Scale ratio (new / old)
+    r = min(new_shape[0] / shape[0], new_shape[1] / shape[1])
+    if not scaleup:  # only scale down, do not scale up (for better test mAP)
+        r = min(r, 1.0)
+
+    # Compute padding
+    ratio = r, r  # width, height ratios
+    new_unpad = int(round(shape[1] * r)), int(round(shape[0] * r))
+    dw, dh = new_shape[1] - new_unpad[0], new_shape[0] - \
+             new_unpad[1]  # wh padding
+    if auto:  # minimum rectangle
+        dw, dh = np.mod(dw, 32), np.mod(dh, 32)  # wh padding
+    elif scaleFill:  # stretch
+        dw, dh = 0.0, 0.0
+        new_unpad = new_shape
+        ratio = new_shape[0] / shape[1], new_shape[1] / \
+                shape[0]  # width, height ratios
+
+    dw /= 2  # divide padding into 2 sides
+    dh /= 2
+
+    if shape[::-1] != new_unpad:  # resize
+        img = cv2.resize(img, new_unpad, interpolation=cv2.INTER_LINEAR)
+    top, bottom = int(round(dh - 0.1)), int(round(dh + 0.1))
+    left, right = int(round(dw - 0.1)), int(round(dw + 0.1))
+    img = cv2.copyMakeBorder(img, top, bottom, left, right,
+                             cv2.BORDER_CONSTANT, value=color)  # add border
+    return img, ratio, (dw, dh)
+
+
+def non_max_suppression(
+        prediction,
+        conf_thres=0.1,
+        iou_thres=0.6,
+        multi_label=True,
+        classes=None,
+        agnostic=False):
+    """
+    Performs  Non-Maximum Suppression on inference results
+    Returns detections with shape:
+        nx6 (x1, y1, x2, y2, conf, cls)
+    """
+
+    # Settings
+    merge = True  # merge for best mAP
+    # (pixels) minimum and maximum box width and height
+    min_wh, max_wh = 2, 4096
+    time_limit = 10.0  # seconds to quit after
+
+    t = time.time()
+    nc = prediction[0].shape[1] - 5  # number of classes
+    multi_label &= nc > 1  # multiple labels per box
+    output = [None] * prediction.shape[0]
+    for xi, x in enumerate(prediction):  # image index, image inference
+        # Apply constraints
+        x = x[x[:, 4] > conf_thres]  # confidence
+        x = x[((x[:, 2:4] > min_wh) & (x[:, 2:4] < max_wh)).all(1)]
+
+        # If none remain process next image
+        if not x.shape[0]:
+            continue
+
+        # Compute conf
+        x[..., 5:] *= x[..., 4:5]  # conf = obj_conf * cls_conf
+
+        # Box (center x, center y, width, height) to (x1, y1, x2, y2)
+        box = xywh2xyxy(x[:, :4])
+
+        # Detections matrix nx6 (xyxy, conf, cls)
+        if multi_label:
+            i, j = (x[:, 5:] > conf_thres).nonzero().t()
+            x = torch.cat((box[i], x[i, j + 5].unsqueeze(1),
+                           j.float().unsqueeze(1)), 1)
+        else:  # best class only
+            conf, j = x[:, 5:].max(1)
+            x = torch.cat(
+                (box, conf.unsqueeze(1), j.float().unsqueeze(1)), 1)[
+                conf > conf_thres]
+
+        # Filter by class
+        if classes:
+            x = x[(j.view(-1, 1) == torch.tensor(classes,
+                                                 device=j.device)).any(1)]
+
+        # Apply finite constraint
+        # if not torch.isfinite(x).all():
+        #     x = x[torch.isfinite(x).all(1)]
+
+        # If none remain process next image
+        n = x.shape[0]  # number of boxes
+        if not n:
+            continue
+
+        # Sort by confidence
+        # x = x[x[:, 4].argsort(descending=True)]
+
+        # Batched NMS
+        c = x[:, 5] * 0 if agnostic else x[:, 5]  # classes
+        boxes, scores = x[:, :4].clone() + c.view(-1, 1) * \
+                        max_wh, x[:, 4]  # boxes (offset by class), scores
+        i = torchvision.ops.boxes.nms(boxes, scores, iou_thres)
+        if merge and (
+                1 < n < 3E3):  # Merge NMS (boxes merged using weighted mean)
+            try:  # update boxes as boxes(i,4) = weights(i,n) * boxes(n,4)
+                iou = box_iou(boxes[i], boxes) > iou_thres  # iou matrix
+                weights = iou * scores[None]  # box weights
+                x[i, :4] = torch.mm(weights, x[:, :4]).float(
+                ) / weights.sum(1, keepdim=True)  # merged boxes
+                # i = i[iou.sum(1) > 1]  # require redundancy
+            except BaseException:
+                # https://github.com/ultralytics/yolov3/issues/1139
+                # print(x, i, x.shape, i.shape)
+                pass
+
+        output[xi] = x[i]
+        if (time.time() - t) > time_limit:
+            break  # time limit exceeded
+
+    return output

yolov3-8.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:caf441af4ff1b82258a7e308a89343f0e5ef9440e89383dde019e030a5b698f2
+size 247866093

yolov3.cfg

@@ -0,0 +1,788 @@
+[net]
+# Testing
+#batch=1
+#subdivisions=1
+# Training
+batch=16
+subdivisions=1
+width=416
+height=416
+channels=3
+momentum=0.9
+decay=0.0005
+angle=0
+saturation = 1.5
+exposure = 1.5
+hue=.1
+
+learning_rate=0.001
+burn_in=1000
+max_batches = 500200
+policy=steps
+steps=400000,450000
+scales=.1,.1
+
+[convolutional]
+batch_normalize=1
+filters=32
+size=3
+stride=1
+pad=1
+activation=leaky
+
+# Downsample
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=3
+stride=2
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=32
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+# Downsample
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=3
+stride=2
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+# Downsample
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=2
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+# Downsample
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=2
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+# Downsample
+
+[convolutional]
+batch_normalize=1
+filters=1024
+size=3
+stride=2
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=1024
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=1024
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=1024
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=1024
+size=3
+stride=1
+pad=1
+activation=leaky
+
+[shortcut]
+from=-3
+activation=linear
+
+######################
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=1024
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=1024
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=1024
+activation=leaky
+
+[convolutional]
+size=1
+stride=1
+pad=1
+filters=255
+activation=linear
+
+
+[yolo]
+mask = 6,7,8
+anchors = 10,13,  16,30,  33,23,  30,61,  62,45,  59,119,  116,90,  156,198,  373,326
+classes=80
+num=9
+jitter=.3
+ignore_thresh = .7
+truth_thresh = 1
+random=1
+
+
+[route]
+layers = -4
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[upsample]
+stride=2
+
+[route]
+layers = -1, 61
+
+
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=512
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=512
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=512
+activation=leaky
+
+[convolutional]
+size=1
+stride=1
+pad=1
+filters=255
+activation=linear
+
+
+[yolo]
+mask = 3,4,5
+anchors = 10,13,  16,30,  33,23,  30,61,  62,45,  59,119,  116,90,  156,198,  373,326
+classes=80
+num=9
+jitter=.3
+ignore_thresh = .7
+truth_thresh = 1
+random=1
+
+
+
+[route]
+layers = -4
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[upsample]
+stride=2
+
+[route]
+layers = -1, 36
+
+
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=256
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=256
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=256
+activation=leaky
+
+[convolutional]
+size=1
+stride=1
+pad=1
+filters=255
+activation=linear
+
+
+[yolo]
+mask = 0,1,2
+anchors = 10,13,  16,30,  33,23,  30,61,  62,45,  59,119,  116,90,  156,198,  373,326
+classes=80
+num=9
+jitter=.3
+ignore_thresh = .7
+truth_thresh = 1
+random=1