Downloads · 30 days
0
m-hamza-mughal/miburi-release-assets
miburi-release-assets is a machine learning model from m-hamza-mughal. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as other.
Demo-time assets for MIBURI: Towards Expressive Interactive Gesture Synthesis (CVPR 2026).
Downloads · 30 days
0
Access
Public
Updated Jun 19, 2026
Repo size
400 MB
Likes
0
Public
Click a slice to open those files.
.npz175 MB · 99%
From the Hugging Face model README
Demo-time assets for MIBURI: Towards Expressive Interactive Gesture Synthesis (CVPR 2026).
These assets are not used standalone — they are loaded by the MIBURI demo. Intended fetch path:
git clone https://github.com/m-hamza-mughal/miburi
cd miburi
pip install .
miburi-download-assets # this repo
miburi-download-checkpoints # https://huggingface.co/m-hamza-mughal/miburi-release-checkpoints
miburi-demo
assets_dep/
demo-static/ # embedded Moshi UI frontend (~1 MB)
mixamo_characters_release/ # 4 Mixamo characters as .npz bundles (~8 MB)
smplx_2020/smplx/ # SMPL-X NEUTRAL 2020 .npz (~115 MB)
The contents of this repository are not MIBURI's to relicense — each file derives from a third-party project under its own terms. By downloading any of these files you agree to honor those upstream licenses.
SMPL-X (smplx_2020/) — Max Planck Society / Perceiving Systems.
Non-commercial research / use only. See the SMPL-X model license:
https://smpl-x.is.tue.mpg.de/modellicense.html.
Citation: Pavlakos et al., Expressive Body Capture: 3D Hands, Face, and
Body from a Single Image, CVPR 2019.
Mixamo characters (mixamo_characters_release/) — Adobe Mixamo.
Subject to the Adobe Mixamo terms of use:
https://helpx.adobe.com/creative-cloud/faq/mixamo-faq.html. The released
.npz bundles repackage Mixamo's free character meshes + skinning;
redistribution here is purely to make the MIBURI demo reproducible. If
you intend to ship Mixamo content in your own product, fetch the source
meshes directly from Mixamo under their terms.
Demo frontend (demo-static/) — derived from Kyutai Moshi's web
frontend (https://github.com/kyutai-labs/moshi); see the upstream Moshi
license (MIT).
The MIBURI code and trained checkpoints (the other half of the demo, not in this repo) are licensed under CC BY-NC 4.0. See the code repository and companion checkpoints repo for the MIBURI-specific terms.
Users of the MIBURI demo additionally need the Moshi weights (Kyutai; fetched automatically from https://huggingface.co/kyutai/moshiko-pytorch-bf16) — those are distributed under CC-BY 4.0 by Kyutai.