MotionGen
AI-style image → motion video demo + backend
Upload images & motion prompt
Upload photos, add a short motion prompt (e.g. "dancing, walking, camera pan") and choose a style. The page will POST files to your backend at /api/generate which will call an AI model to return a generated video URL.
Paste your backend endpoint here. The frontend will POST multipart/form-data containing images and the prompt.
Tip: The backend examples (Node.js + Python worker) are included in README below in this canvas. The backend runs the heavy AI model — the frontend just uploads images and shows returned video URLs.
Backend starter (Node.js + Python worker)
README Files included in this starter (paste into your server project): 1) server.js — Express server that accepts multipart uploads and enqueues a job (calls python worker or forwards to remote API). 2) worker/generate.py — Python worker script (placeholder) showing how to call a model or external API to generate video from an image + prompt. You must install and run a model or call a hosted API. 3) package.json — Node dependencies. 4) Dockerfile — Basic container for the Node server. ==== server.js ==== // Paste server.js content into your project (see canvas file) — this endpoint saves uploads to /tmp/uploads and invokes the worker. ==== worker/generate.py ==== # Python script placeholder: replace with your model invocation (AnimateDiff, StableVideo, Pika API, Runway)
Comments
Post a Comment