Upload utils.py

src/spin/utils.py

@@ -0,0 +1,146 @@
+import json
+
+import cv2
+import numpy as np
+import torch
+from skimage.transform import resize, rotate
+from torchvision.transforms import Normalize
+
+from .constants import IMG_NORM_MEAN, IMG_NORM_STD, IMG_RES
+
+
+def get_transform(center, scale, res, rot=0):
+    """Generate transformation matrix."""
+    h = 200 * scale
+    t = np.zeros((3, 3))
+    t[0, 0] = float(res[1]) / h
+    t[1, 1] = float(res[0]) / h
+    t[0, 2] = res[1] * (-float(center[0]) / h + 0.5)
+    t[1, 2] = res[0] * (-float(center[1]) / h + 0.5)
+    t[2, 2] = 1
+    if not rot == 0:
+        rot = -rot  # To match direction of rotation from cropping
+        rot_mat = np.zeros((3, 3))
+        rot_rad = rot * np.pi / 180
+        sn, cs = np.sin(rot_rad), np.cos(rot_rad)
+        rot_mat[0, :2] = [cs, -sn]
+        rot_mat[1, :2] = [sn, cs]
+        rot_mat[2, 2] = 1
+        # Need to rotate around center
+        t_mat = np.eye(3)
+        t_mat[0, 2] = -res[1] / 2
+        t_mat[1, 2] = -res[0] / 2
+        t_inv = t_mat.copy()
+        t_inv[:2, 2] *= -1
+        t = np.dot(t_inv, np.dot(rot_mat, np.dot(t_mat, t)))
+
+    return t
+
+
+def transform(pt, center, scale, res, invert=0, rot=0):
+    """Transform pixel location to different reference."""
+    t = get_transform(center, scale, res, rot=rot)
+    if invert:
+        t = np.linalg.inv(t)
+    new_pt = np.array([pt[0] - 1, pt[1] - 1, 1.0]).T
+    new_pt = np.dot(t, new_pt)
+
+    return new_pt[:2].astype(int) + 1
+
+
+def crop(img, center, scale, res, rot=0):
+    """Crop image according to the supplied bounding box."""
+    # Upper left point
+    ul = np.array(transform([1, 1], center, scale, res, invert=1)) - 1
+    # Bottom right point
+    br = np.array(transform([res[0] + 1, res[1] + 1], center, scale, res, invert=1)) - 1
+
+    # Padding so that when rotated proper amount of context is included
+    pad = int(np.linalg.norm(br - ul) / 2 - float(br[1] - ul[1]) / 2)
+    if not rot == 0:
+        ul -= pad
+        br += pad
+
+    new_shape = [br[1] - ul[1], br[0] - ul[0]]
+    if len(img.shape) > 2:
+        new_shape += [img.shape[2]]
+    new_img = np.zeros(new_shape)
+
+    # Range to fill new array
+    new_x = max(0, -ul[0]), min(br[0], len(img[0])) - ul[0]
+    new_y = max(0, -ul[1]), min(br[1], len(img)) - ul[1]
+    # Range to sample from original image
+    old_x = max(0, ul[0]), min(len(img[0]), br[0])
+    old_y = max(0, ul[1]), min(len(img), br[1])
+    new_img[new_y[0] : new_y[1], new_x[0] : new_x[1]] = img[
+        old_y[0] : old_y[1], old_x[0] : old_x[1]
+    ]
+
+    if not rot == 0:
+        # Remove padding
+        new_img = rotate(new_img, rot)
+        new_img = new_img[pad:-pad, pad:-pad]
+
+    new_img = resize(new_img, res)
+
+    return new_img
+
+
+def bbox_from_openpose(openpose_file, rescale=1.2, detection_thresh=0.2):
+    """Get center and scale for bounding box from openpose detections."""
+    with open(openpose_file, "r") as f:
+        keypoints = json.load(f)["people"][0]["pose_keypoints_2d"]
+    keypoints = np.reshape(np.array(keypoints), (-1, 3))
+    valid = keypoints[:, -1] > detection_thresh
+    valid_keypoints = keypoints[valid][:, :-1]
+    center = valid_keypoints.mean(axis=0)
+    bbox_size = (valid_keypoints.max(axis=0) - valid_keypoints.min(axis=0)).max()
+    # adjust bounding box tightness
+    scale = bbox_size / 200.0
+    scale *= rescale
+
+    return center, scale
+
+
+def bbox_from_json(bbox_file):
+    """Get center and scale of bounding box from bounding box annotations.
+    The expected format is [top_left(x), top_left(y), width, height].
+    """
+    with open(bbox_file, "r") as f:
+        bbox = np.array(json.load(f)["bbox"]).astype(np.float32)
+    ul_corner = bbox[:2]
+    center = ul_corner + 0.5 * bbox[2:]
+    width = max(bbox[2], bbox[3])
+    scale = width / 200.0
+    # make sure the bounding box is rectangular
+    return center, scale
+
+
+def process_image(img_file, bbox_file=None, openpose_file=None, input_res=IMG_RES):
+    """Read image, do preprocessing and possibly crop it according to the bounding box.
+    If there are bounding box annotations, use them to crop the image.
+    If no bounding box is specified but openpose detections are available, use them to get the bounding box.
+    """
+    img_file = str(img_file)
+    normalize_img = Normalize(mean=IMG_NORM_MEAN, std=IMG_NORM_STD)
+    img = cv2.imread(img_file)[
+        :, :, ::-1
+    ].copy()  # PyTorch does not support negative stride at the moment
+    if bbox_file is None and openpose_file is None:
+        # Assume that the person is centerered in the image
+        height = img.shape[0]
+        width = img.shape[1]
+        center = np.array([width // 2, height // 2])
+        scale = max(height, width) / 200
+    else:
+        if bbox_file is not None:
+            center, scale = bbox_from_json(bbox_file)
+        elif openpose_file is not None:
+            center, scale = bbox_from_openpose(openpose_file)
+
+    img = crop(img, center, scale, (input_res, input_res))
+    img = img.astype(np.float32) / 255.0
+    img = torch.from_numpy(img).permute(2, 0, 1)
+    norm_img = normalize_img(img.clone())
+
+    return img, norm_img