MCG-NJU/SteadyDancer-GGUF

SteadyDancer: Harmonized and Coherent Human Image Animation with First-Frame Preservation

Jiaming Zhang · Shengming Cao · Rui Li · Xiaotong Zhao · Yutao Cui
Xinglin Hou · Gangshan Wu · Haolan Chen · Yu Xu · Limin Wang · Kai Ma


Multimedia Computing Group, Nanjing University   |   Platform and Content Group (PCG), Tencent

• This repository is the checkpoint of paper "SteadyDancer: Harmonized and Coherent Human Image Animation with First-Frame Preservation" in GGUF format, which coverted from Kijai weight by ComfyUI-GGUF. It offers lower cost in Kijai ComfyUI.
• If you encounter problem RuntimeError: expected scalar type Float but found BFloat16, please pull the latest code of Kijai ComfyUI.
• For each quantized version, we provide two weights: *_fix_5d_tensor.gguf indicating the convolutional weights come from Wan21_I2V_SteadyDancer_fp16.safetensors (i.e., fp16 format); and *_fix_5d_tensor_from_fp8_e4m3fn_scaled_KJ.gguf indicating the convolutional weights come from Wan2_1-I2V-14B-480p_fp8_e4m3fn_scaled_KJ.safetensors (i.e., fp32 format). See https://github.com/kijai/ComfyUI-WanVideoWrapper/issues/1702#issuecomment-3611568304 for details. Both weights are available and the results are largely consistent. We are working to determine the final version that will remain.

📚 Citation

If you find our paper or this codebase useful for your research, please cite us.

@misc{zhang2025steadydancer,
      title={SteadyDancer: Harmonized and Coherent Human Image Animation with First-Frame Preservation}, 
      author={Jiaming Zhang and Shengming Cao and Rui Li and Xiaotong Zhao and Yutao Cui and Xinglin Hou and Gangshan Wu and Haolan Chen and Yu Xu and Limin Wang and Kai Ma},
      year={2025},
      eprint={2511.19320},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2511.19320}, 
}