I get exactly what you are doing but I don't understand what would be novel about it? It really looks just like a simple SD/wav2vec/insightface/animatediff pipe

I get exactly what you are doing but I don't understand what would be novel about it? It really looks just like a simple SD/wav2vec/insightface/animatediff pipeline, everyone can plug together in comfy or with diffusers. The muscle claim is also a bit dubious... PS: With insightface models in your reqs the oss aspect is also pretty much void for any use other than research and your readme should reflect that.