File size: 6,252 Bytes
9a2edf3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
import cv2
import torch
import numpy as np
from PIL import Image
import torchvision.transforms as transforms
from ultralytics import YOLO
import time
import os
import tempfile
from flask import Flask, request, jsonify
import gradio as gr

# Initialize Flask app and Gradio interface
app = Flask(__name__)

# Global variable to store detection history
detection_history = []

# Emotion labels
emotions = ['Angry', 'Disgust', 'Fear', 'Happy', 'Sad', 'Surprise', 'Neutral']

# Load models (cache in Hugging Face Space)
def load_models():
    # Face detection model
    face_model = YOLO('yolov8n-face.pt')
    
    # Emotion model (simplified version of your CNN)
    class EmotionCNN(torch.nn.Module):
        def __init__(self, num_classes=7):
            super().__init__()
            self.features = torch.nn.Sequential(
                torch.nn.Conv2d(1, 64, 3, padding=1),
                torch.nn.ReLU(),
                torch.nn.MaxPool2d(2),
                torch.nn.Conv2d(64, 128, 3, padding=1),
                torch.nn.ReLU(),
                torch.nn.MaxPool2d(2),
                torch.nn.Conv2d(128, 256, 3, padding=1),
                torch.nn.ReLU(),
                torch.nn.MaxPool2d(2)
            )
            self.classifier = torch.nn.Sequential(
                torch.nn.Dropout(0.5),
                torch.nn.Linear(256*6*6, 1024),
                torch.nn.ReLU(),
                torch.nn.Dropout(0.5),
                torch.nn.Linear(1024, num_classes)
            )
        
        def forward(self, x):
            x = self.features(x)
            x = torch.flatten(x, 1)
            x = self.classifier(x)
            return x
    
    emotion_model = EmotionCNN()
    # Load your pretrained weights here
    # emotion_model.load_state_dict(torch.load('emotion_model.pth'))
    emotion_model.eval()
    
    return face_model, emotion_model

face_model, emotion_model = load_models()

# Preprocessing function
def preprocess_face(face_img):
    transform = transforms.Compose([
        transforms.Resize((48, 48)),
        transforms.Grayscale(),
        transforms.ToTensor(),
        transforms.Normalize(mean=[0.5], std=[0.5])
    ])
    face_pil = Image.fromarray(cv2.cvtColor(face_img, cv2.COLOR_BGR2RGB))
    return transform(face_pil).unsqueeze(0)

# Process video function
def process_video(video_path):
    global detection_history
    detection_history = []
    
    cap = cv2.VideoCapture(video_path)
    if not cap.isOpened():
        return {"error": "Could not open video"}
    
    frame_count = 0
    fps = cap.get(cv2.CAP_PROP_FPS)
    frame_skip = int(fps / 3)  # Process ~3 frames per second
    
    while True:
        ret, frame = cap.read()
        if not ret:
            break
            
        frame_count += 1
        if frame_count % frame_skip != 0:
            continue
            
        # Face detection
        results = face_model(frame)
        
        for result in results:
            boxes = result.boxes
            if len(boxes) == 0:
                continue
                
            for box in boxes:
                x1, y1, x2, y2 = map(int, box.xyxy[0].tolist())
                face_img = frame[y1:y2, x1:x2]
                
                if face_img.size == 0:
                    continue
                    
                # Emotion prediction
                face_tensor = preprocess_face(face_img)
                with torch.no_grad():
                    output = emotion_model(face_tensor)
                    prob = torch.nn.functional.softmax(output, dim=1)[0]
                    pred_idx = torch.argmax(output).item()
                    confidence = prob[pred_idx].item()
                
                detection_history.append({
                    "frame": frame_count,
                    "time": frame_count / fps,
                    "emotion": emotions[pred_idx],
                    "confidence": confidence,
                    "box": [x1, y1, x2, y2]
                })
    
    cap.release()
    
    if not detection_history:
        return {"error": "No faces detected"}
    
    return {
        "detections": detection_history,
        "summary": {
            "total_frames": frame_count,
            "fps": fps,
            "duration": frame_count / fps
        }
    }

# Flask API endpoint
@app.route('/api/predict', methods=['POST'])
def api_predict():
    if 'file' not in request.files:
        return jsonify({"error": "No file provided"}), 400
        
    file = request.files['file']
    if file.filename == '':
        return jsonify({"error": "No selected file"}), 400
    
    # Save to temp file
    temp_path = os.path.join(tempfile.gettempdir(), file.filename)
    file.save(temp_path)
    
    # Process video
    result = process_video(temp_path)
    
    # Clean up
    os.remove(temp_path)
    
    return jsonify(result)

# Gradio interface
def gradio_predict(video):
    temp_path = os.path.join(tempfile.gettempdir(), video.name)
    with open(temp_path, 'wb') as f:
        f.write(video.read())
    
    result = process_video(temp_path)
    os.remove(temp_path)
    
    if "error" in result:
        return result["error"]
    
    # Create visualization
    cap = cv2.VideoCapture(video.name)
    ret, frame = cap.read()
    cap.release()
    
    if ret:
        # Draw last detection on frame
        last_det = result["detections"][-1]
        x1, y1, x2, y2 = last_det["box"]
        cv2.rectangle(frame, (x1, y1), (x2, y2), (0, 255, 0), 2)
        cv2.putText(frame, f"{last_det['emotion']} ({last_det['confidence']:.2f})", 
                   (x1, y1-10), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (0, 255, 0), 2)
        
        # Convert to RGB for Gradio
        frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
        return frame, result
    return result

# Create Gradio interface
demo = gr.Interface(
    fn=gradio_predict,
    inputs=gr.Video(label="Upload Video"),
    outputs=[
        gr.Image(label="Detection Preview"), 
        gr.JSON(label="Results")
    ],
    title="Video Emotion Detection",
    description="Upload a video to detect emotions in faces"
)

# Mount Gradio app
app = gr.mount_gradio_app(app, demo, path="/")

if __name__ == "__main__":
    app.run(host="0.0.0.0", port=7860)