ReShot — copy the shot, not the actors. Video → depth map, OpenPose skeleton or canny lines for Seedance 2.0/2.5 reference video, MiniMax H3 Fun ControlNet and Wan VACE. Apache-2.0. Open-sourced by Maosika 猫斯卡 (www.maosika.com), the AI short-drama production system.
Contributing
Thanks for looking. Small, focused PRs are easiest to review.
Setup
git clone [email protected]:maosika-ai/reshot.git && cd reshot
python -m venv .venv && . .venv/bin/activate
pip install torch torchvision # pick the build for your machine
pip install -e ".[ffmpeg,dev]"
Before you push
ruff check reshot tests && ruff format reshot tests
pytest -q # ~30 tests, no model, ~2 s
RUN_MODEL_TESTS=1 pytest -q tests/test_model_smoke.py # optional: real model on CPU, downloads 111 MB
CI runs the same on Ubuntu and Windows, plus the model smoke on CPU.
Ground rules
reshot/third_party/is vendored upstream code — do not edit it; fix things in our layer.- User-facing failures raise a
ReshotErrorsubclass with a concrete fix in the message. - Anything that changes output (normalisation, encoding, presets) needs a test and a CHANGELOG line, and the README/USAGE tables updated in both languages.
- Performance claims in docs must come from a measurement; say where it was measured.
Reporting a bug
Include: OS, GPU/CPU, python -c "import torch;print(torch.__version__)", the exact command,
and the stderr output with -v. If a clip reproduces it and you can share it, link it.
Where to talk
Questions and results go to Discussions; bugs go to issues. If you want to take on something larger than a good-first-issue (pose output, a new backend), open a thread in Ideas first so we can agree on the shape before you build.