app.py

import gradio as gr
import os
import cv2
import numpy as np
from moviepy.editor import *

token = os.environ.get('HF_TOKEN')
pix2pix = gr.Blocks.load(name="spaces/fffiloni/instruct-pix2pix-clone", api_key=token)

def get_frames(video_in):
    frames = []
    # Opens the Video file
    cap= cv2.VideoCapture(video_in)
    fps = cap.get(cv2.CAP_PROP_FPS)
    i=0
    while(cap.isOpened()):
        ret, frame = cap.read()
        if ret == False:
            break
        cv2.imwrite('kang'+str(i)+'.jpg',frame)
        frames.append('kang'+str(i)+'.jpg')
        i+=1
    
    cap.release()
    cv2.destroyAllWindows()

    return frames, fps

def create_video(frames, fps):
    
    clip = ImageSequenceClip(frames, fps=fps)
    clip.write_videofile("movie.mp4", fps=fps)
    
    return 'movie.mp4'


def infer(prompt,video_in, seed_in):
    break_vid = get_frames(video_in)
    
    frames_list= break_vid[0]
    fps = break_vid[1]
    result_frames = []
    
    for i in frames_list:
        pix2pix_img = pix2pix(prompt,5.5,1.5,i,15,"",512,512,seed_in,fn_index=0)
        images = [os.path.join(pix2pix_img[0], img) for img in os.listdir(pix2pix_img[0])]
        result_frames.append(images[0])

    final_vid = create_video(result_frames, fps)
    
    return final_vid


with gr.Blocks(css='style.css') as demo:
    with gr.Column(elem_id="col-container"):
        with gr.Row():
            with gr.Row():
                prompt = gr.Textbox(placeholder="enter prompt")
                video_inp = gr.Video(label="Video source", source="upload", type="filepath")
                seed_inp = gr.Number(label="Seed", value=123456)
            with gr.Column():
                video_out = gr.Video(label="Pix2pix video result")
                submit_btn = gr.Button("Generate Pix2Pix video")

    inputs = [prompt,video_inp,seed_inp]
    outputs = [video_out]
    
    submit_btn.click(infer, inputs, outputs)

demo.launch().queue(max_size=12)