Video Enhancing Recipes

import requests
import time

API_KEY = "YOUR API KEY"
BASE_URL = "https://backend.tensorpix.ai/api"
STATUS_CHECK_INTERVAL = 10
HEADERS = {"Authorization": f"Token {API_KEY}"}

def start_video_enhancement(video_url: str, ml_models: list[int]) -> dict:
    data = {
        "codec": "libx264",
        "crf": 23,
        "container": "mp4",
        "chroma_subsampling": "yuv420p",
        "video_url": video_url,
        "ml_models": ml_models,
        "output_resolution": -1,
        "grain": 0,
        "stabilization_smoothing": 0,
    }
    response = requests.post(url=f"{BASE_URL}/jobs/from-url/", headers=HEADERS, data=data)
    response.raise_for_status()
    return response.json()

def wait_for_job_finish(job_id: int) -> dict:
    while True:
        time.sleep(STATUS_CHECK_INTERVAL)
        job_data = requests.get(url=f"{BASE_URL}/jobs/{job_id}/", headers=HEADERS).json()
        status = job_data["status"]
        if status == 0:
            print("Job is in queue")
        elif status == 1:
            print(f"Progress {(job_data['processing_progress'] * 100):.2f}%")
        elif status == 2:
            return job_data
        else:
            raise Exception("Job has failed")

def remove_video(video_id):
    requests.delete(url=f"{BASE_URL}/restored-videos/{video_id}/", headers=HEADERS)

job_info = start_video_enhancement(
    video_url="https://cdn.tensorpix.ai/videos/landing_hero_v2.mp4",
    ml_models=[34] # List all models with `/api/ml-models/` endpoint
)
print(job_info)

completed_job_data = wait_for_job_finish(job_info["id"])

# Getting enhanced video information
enhanced_video_url = completed_job_data["output_video"]["file"]
print(f"Enhanced video available at: {enhanced_video_url}")

# Cleaning up -- removing the enhanced video from TensorPix servers
# remove_video(completed_job_data["output_video"]["id"])