Improve paper and artifact metadata
Browse files
README.md
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: other
|
| 3 |
+
library_name: pytorch
|
| 4 |
+
tags:
|
| 5 |
+
- co-speech-gesture-generation
|
| 6 |
+
- co-speech-motion-generation
|
| 7 |
+
- speech-driven-motion
|
| 8 |
+
- diffusion-models
|
| 9 |
+
- long-horizon-generation
|
| 10 |
+
- beat2
|
| 11 |
+
- smpl-x
|
| 12 |
+
- globaldiff
|
| 13 |
+
datasets:
|
| 14 |
+
- H-Liu1997/BEAT2
|
| 15 |
+
---
|
| 16 |
+
|
| 17 |
+
# GlobalDiff official checkpoints
|
| 18 |
+
|
| 19 |
+
Official checkpoints for **Mitigating Error Accumulation in Co-Speech Motion Generation via Global Rotation Diffusion and Multi-Level Constraints** (AAAI 2026), a diffusion approach to long-horizon **co-speech gesture generation**.
|
| 20 |
+
|
| 21 |
+
- [Paper](https://arxiv.org/abs/2511.10076) 路 [Hugging Face Paper page](https://huggingface.co/papers/2511.10076)
|
| 22 |
+
- [Project page](https://xiangyuezhang.com/GlobalDiff/) 路 [Code](https://github.com/Xiangyue-Zhang/GlobalDiff)
|
| 23 |
+
- [Published version](https://doi.org/10.1609/aaai.v40i15.38281)
|
| 24 |
+
|
| 25 |
+
## Files
|
| 26 |
+
|
| 27 |
+
The checkpoint files mirror the official repository while preserving their repository-relative paths:
|
| 28 |
+
|
| 29 |
+
| Path | Role |
|
| 30 |
+
| --- | --- |
|
| 31 |
+
| `Scripts/VAE/ckpt/split/global/MaskedVAE2-HorizonFlip/best.pt` | Global rotation VAE checkpoint |
|
| 32 |
+
| `Scripts/VAE/ckpt/split/global/MaskedVAE3-HorizonFlip/best.pt` | Global rotation VAE checkpoint |
|
| 33 |
+
| `Scripts/FM/ckpt/split/SimpleSpeechModel/best.zip` | Speech model checkpoint archive |
|
| 34 |
+
| `Scripts/FM/EMAGE_VAE/AESKConv_240_100.bin` | Evaluation/representation asset |
|
| 35 |
+
| `best_pid_2.zip` | Released Speaker 2 generated test results |
|
| 36 |
+
|
| 37 |
+
See the code repository's environment checker and inference instructions for all runtime dependencies.
|
| 38 |
+
|
| 39 |
+
## Terms
|
| 40 |
+
|
| 41 |
+
This repository mirrors research artifacts released by the authors. No new license is granted by this model card. Code, BEAT2, SMPL-X, pretrained encoders, and other third-party assets remain subject to their respective terms.
|
| 42 |
+
|
| 43 |
+
## Citation
|
| 44 |
+
|
| 45 |
+
```bibtex
|
| 46 |
+
@inproceedings{zhang2026globaldiff,
|
| 47 |
+
title={Mitigating Error Accumulation in Co-Speech Motion Generation via Global Rotation Diffusion and Multi-Level Constraints},
|
| 48 |
+
author={Zhang, Xiangyue and Li, Jianfang and Ren, Jianqiang and Zhang, Jiaxu},
|
| 49 |
+
booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
|
| 50 |
+
volume={40},
|
| 51 |
+
number={15},
|
| 52 |
+
pages={12834--12842},
|
| 53 |
+
year={2026},
|
| 54 |
+
doi={10.1609/aaai.v40i15.38281}
|
| 55 |
+
}
|
| 56 |
+
```
|