Spaces:
Sleeping
Sleeping
File size: 5,027 Bytes
f8182db 99e245d d74ebd5 a51300a d328d4c 60b5a4c d7454ed 55d1370 c4fa864 56666a2 949b93c f8182db d24bd0f d7454ed 9f2a1cd d7454ed 9f2a1cd d7454ed 0f0d8fb d7454ed e28fcd2 d7454ed d24bd0f d7454ed d24bd0f 09213d6 d24bd0f d74ebd5 d24bd0f ce6d9ab d24bd0f ce6d9ab d24bd0f d7454ed f8182db ed609ce d24bd0f e28fcd2 2491068 d7454ed 2491068 f8182db 068c371 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 |
import gradio as gr
import os
import cv2
from encoded_video import EncodedVideo, write_video
import torch
import numpy as np
from torchvision.datasets import ImageFolder
from transformers import ViTFeatureExtractor, ViTForImageClassification, AutoFeatureExtractor, ViTMSNForImageClassification
from pathlib import Path
import pytorch_lightning as pl
from torch.utils.data import DataLoader
from torchmetrics import Accuracy
def video_identity(video,user_name,class_name,trainortest,ready):
if ready=='yes':
data_dir = Path(str(user_name)+'/train')
train_ds = ImageFolder(data_dir)
test_dir = Path(str(user_name)+'/test')
test_ds = ImageFolder(test_dir)
label2id = {}
id2label = {}
for i, class_name in enumerate(train_ds.classes):
label2id[class_name] = str(i)
id2label[str(i)] = class_name
class ImageClassificationCollator:
def __init__(self, feature_extractor):
self.feature_extractor = feature_extractor
def __call__(self, batch):
encodings = self.feature_extractor([x[0] for x in batch], return_tensors='pt')
encodings['labels'] = torch.tensor([x[1] for x in batch], dtype=torch.long)
return encodings
feature_extractor = ViTFeatureExtractor.from_pretrained('google/vit-base-patch16-224-in21k')
model = ViTForImageClassification.from_pretrained(
'google/vit-base-patch16-224-in21k',
num_labels=len(label2id),
label2id=label2id,
id2label=id2label
)
collator = ImageClassificationCollator(feature_extractor)
class Classifier(pl.LightningModule):
def __init__(self, model, lr: float = 2e-5, **kwargs):
super().__init__()
self.save_hyperparameters('lr', *list(kwargs))
self.model = model
self.forward = self.model.forward
self.val_acc = Accuracy(
task='multiclass' if model.config.num_labels > 2 else 'binary',
num_classes=model.config.num_labels
)
def training_step(self, batch, batch_idx):
outputs = self(**batch)
self.log(f"train_loss", outputs.loss)
return outputs.loss
def validation_step(self, batch, batch_idx):
outputs = self(**batch)
self.log(f"val_loss", outputs.loss)
acc = self.val_acc(outputs.logits.argmax(1), batch['labels'])
self.log(f"val_acc", acc, prog_bar=True)
return outputs.loss
def configure_optimizers(self):
return torch.optim.Adam(self.parameters(), lr=self.hparams.lr)
train_loader = DataLoader(train_ds, batch_size=8, collate_fn=collator, num_workers=8, shuffle=True)
test_loader = DataLoader(test_ds, batch_size=8, collate_fn=collator, num_workers=2)
for name, param in model.named_parameters():
param.requires_grad = False
if name.startswith("classifier"): # choose whatever you like here
param.requires_grad = True
pl.seed_everything(42)
classifier = Classifier(model, lr=2e-5)
trainer = pl.Trainer(accelerator='cpu', devices=1, precision=16, max_epochs=3)
trainer.fit(classifier, train_loader, test_loader)
for batch_idx, data in enumerate(test_loader):
outputs = model(**data)
img=data['pixel_values'][0][0]
preds=str(outputs.logits.softmax(1).argmax(1))
labels=str(data['labels'])
return img, preds, labels
else:
capture = cv2.VideoCapture(video)
user_d=str(user_name)+'/'+str(trainortest)
class_d=str(user_name)+'/'+str(trainortest)+'/'+str(class_name)
if not os.path.exists(user_d):
os.makedirs(user_d)
if not os.path.exists(class_d):
os.makedirs(class_d)
frameNr = 0
while (True):
success, frame = capture.read()
if success:
cv2.imwrite(f'{class_d}/frame_{frameNr}.jpg', frame)
else:
break
frameNr = frameNr+10
img=cv2.imread(class_d+'/frame_0.jpg')
return img, trainortest, class_d
demo = gr.Interface(video_identity,
inputs=[gr.Video(source='upload'),
gr.Text(),
gr.Text(),
gr.Text(label='Which set is this? (type train or test)'),
gr.Text(label='Are you ready? (type yes or no)')],
outputs=[gr.Image(),
gr.Text(),
gr.Text()],
cache_examples=True)
demo.launch(debug=True) |