app.py

import torch
import cv2
import numpy as np
import gradio as gr
from ultralytics import YOLO

# Check if CUDA (GPU support) is available
# device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

# Load the YOLOv8 model
model = YOLO('yolov8x-seg.pt')

def process_video(input_video_path):
    cap = cv2.VideoCapture(input_video_path)
    if not cap.isOpened():
        print("Error: Couldn't open the video file.")
        return

    frame_width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
    frame_height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
    fps = cap.get(cv2.CAP_PROP_FPS)
    fourcc = cv2.VideoWriter_fourcc(*'mp4v')
    out = cv2.VideoWriter("output_video.mp4", fourcc, fps, (frame_width, frame_height))

    while cap.isOpened():
        ret, frame = cap.read()
        if not ret:
            break

        # Resize frame to match the expected input shape of the model
        resized_frame = cv2.resize(frame, (640, 640))

        # Convert resized frame to torch tensor and move it to GPU
        frame_tensor = torch.from_numpy(resized_frame).permute(2, 0, 1).unsqueeze(0).float() / 255.0

        threshold = 0.2
        frame_copy = frame.copy()
        results = model(frame_tensor)[0]
        for result in results.boxes.data.tolist():
            x1, y1, x2, y2, score, class_id = result
            if score > threshold:
                cv2.rectangle(frame_copy, (int(x1), int(y1)), (int(x2), int(y2)), (0, 255, 0), 4)
                cv2.putText(frame_copy, results.names[int(class_id)].upper(), (int(x1), int(y1 - 10)),
                            cv2.FONT_HERSHEY_SIMPLEX, 1.3, (0, 255, 0), 3, cv2.LINE_AA)
                cv2.putText(frame_copy, str(score), (int(x1), int(y2 + 10)),
                            cv2.FONT_HERSHEY_SIMPLEX, 1.3, (0, 0, 255), 3, cv2.LINE_AA)
        out.write(frame_copy)

    cap.release()
    out.release()
    cv2.destroyAllWindows()

    return "output_video.mp4"

# Define the input and output interfaces for Gradio
inputs_video = gr.Video(label="Input Video")
outputs_video = gr.Video(label="Output Video")

# Create the Gradio interface
demo = gr.Interface(
    fn=process_video,
    inputs=inputs_video,
    outputs=outputs_video,
    title="Animal detector using YOLOv8 NANO for Videos (GPU)",
)

# Launch the interface
demo.launch()