app.py

import os
import cv2
import torch
import numpy as np
from PIL import Image
import torchvision.transforms as transforms
import time
import json
from typing import Dict, Any
from fastapi import FastAPI, HTTPException, File, UploadFile
from pydantic import BaseModel
import gradio as gr
import shutil
import tempfile

app = FastAPI()

# Global variable to store the history of largest face detections
largest_face_detections = []

# EmotionCNN model definition (same as in your original code)
class EmotionCNN(torch.nn.Module):
    def __init__(self, num_classes=7):
        super(EmotionCNN, self).__init__()
        
        # Your convolutional layers and other definitions
        # ...

    def forward(self, x):
        # Forward method as in your code
        pass

# Load emotion model
def load_emotion_model(model_path, device='cuda' if torch.cuda.is_available() else 'cpu'):
    checkpoint = torch.load(model_path, map_location=device)
    model = EmotionCNN(num_classes=7)
    model.load_state_dict(checkpoint['model_state_dict'])
    model.to(device)
    model.eval()
    return model

# Process the uploaded video (either MP4 or WebM)
def process_video(video_file: UploadFile) -> Dict[str, Any]:
    global largest_face_detections
    largest_face_detections = []  # Reset detections for new video
    
    # Path to models and other setup
    face_cascade_path = cv2.data.haarcascades + 'haarcascade_frontalface_default.xml'
    emotion_model_path = "best_emotion_model.pth"
    
    if not os.path.exists(face_cascade_path):
        raise HTTPException(status_code=400, detail="Face cascade classifier not found")
        
    if not os.path.exists(emotion_model_path):
        raise HTTPException(status_code=400, detail="Emotion model not found")
    
    device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
    
    try:
        face_cascade = cv2.CascadeClassifier(face_cascade_path)
        emotion_model = load_emotion_model(emotion_model_path, device)
    except Exception as e:
        raise HTTPException(status_code=500, detail=f"Error loading models: {str(e)}")
        
    emotions = ['Angry', 'Disgust', 'Fear', 'Happy', 'Sad', 'Surprise', 'Neutral']
    
    # Save the uploaded video file to a temporary directory
    temp_dir = tempfile.mkdtemp()
    video_path = os.path.join(temp_dir, "uploaded_video")
    with open(video_path, "wb") as buffer:
        shutil.copyfileobj(video_file.file, buffer)
    
    cap = cv2.VideoCapture(video_path)
    if not cap.isOpened():
        raise HTTPException(status_code=400, detail=f"Could not open video file at {video_path}")
    
    frame_count = 0
    total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
    
    while True:
        ret, frame = cap.read()
        if not ret:
            break
            
        frame_count += 1
        
        largest_face_area = 0
        current_detection = None
        
        gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
        
        faces = face_cascade.detectMultiScale(gray, scaleFactor=1.1, minNeighbors=5, minSize=(30, 30))
        
        for (x, y, w, h) in faces:
            face_area = w * h
            margin = 20
            x1 = max(0, x - margin)
            y1 = max(0, y - margin)
            x2 = min(frame.shape[1], x + w + margin)
            y2 = min(frame.shape[0], y + h + margin)
            
            face_img = frame[y1:y2, x1:x2]
            
            if face_img.size == 0 or face_img.shape[0] < 20 or face_img.shape[1] < 20:
                continue
            
            face_tensor = preprocess_face(face_img)
            
            with torch.no_grad():
                face_tensor = face_tensor.to(device)
                output = emotion_model(face_tensor)
                probabilities = torch.nn.functional.softmax(output, dim=1)
                emotion_idx = torch.argmax(output, dim=1).item()
                confidence = probabilities[0][emotion_idx].item()
              
            emotion = emotions[emotion_idx]
            
            if face_area > largest_face_area:
                largest_face_area = face_area
                current_detection = {
                    'emotion': emotion,
                    'confidence': confidence,
                    'timestamp': time.time(),
                    'frame_number': frame_count
                }
        
        if current_detection:
            largest_face_detections.append(current_detection)
    
    cap.release()
    
    if not largest_face_detections:
        return {
            "success": True,
            "message": "No faces detected in video",
            "results": [],
            "error": None
        }
    
    emotions_count = {}
    for detection in largest_face_detections:
        emotion = detection['emotion']
        emotions_count[emotion] = emotions_count.get(emotion, 0) + 1
    
    dominant_emotion = max(emotions_count.items(), key=lambda x: x[1])[0]
    
    return {
        "success": True,
        "message": "Video processed successfully",
        "results": {
            "detections": largest_face_detections,
            "summary": {
                "total_frames": total_frames,
                "total_detections": len(largest_face_detections),
                "emotions_count": emotions_count,
                "dominant_emotion": dominant_emotion
            }
        },
        "error": None
    }

class VideoRequest(BaseModel):
    path: str

# FastAPI endpoint for processing the video file
@app.post("/process_video/")
async def process_video_request(file: UploadFile = File(...)):
    try:
        results = process_video(file)
        return results
    except Exception as e:
        raise HTTPException(status_code=500, detail=str(e))

# Gradio interface
def gradio_interface():
    def process_gradio_video(video_file):
        # This function now accepts WebM files and other video formats.
        return process_video(video_file)

    interface = gr.Interface(
        fn=process_gradio_video, 
        inputs=gr.inputs.Video(type="file"),  # 'file' ensures that Gradio handles all formats including WebM
        outputs="json"
    )

    return interface

# Launch Gradio Interface on FastAPI
gradio_interface().launch(server_name="0.0.0.0", server_port=7860, share=True)