OpenTalker/SadTalker ? reverse-engineered prompt

Build me a simple local app that turns one portrait photo and a voice clip into a realistic talking face video. I want an easy browser UI where I can upload an image and audio file, pick a few simple options like still mode, full image mode, resize, and face enhancement, then generate, preview, and download the result.

Please make the setup as painless as possible, including getting any required model files in place and using sensible defaults so it works out of the box. If this repo already has a demo or launcher, use that and fix whatever is missing so a normal person can run it without digging through a bunch of docs. A basic command line option would be great too for quick runs.

Make sure it accepts common image and audio formats, saves outputs in a clear folder, and shows friendly errors if models are missing or video tools are not installed. Look up the current docs online if you need to, but keep the final experience simple.