Release artifacts (EmoCAST model linking, dataset) on Hugging Face

[NielsRogge]

Hi @Jiang-maomao 🤗

Niels here from the open-source team at Hugging Face. I discovered your work on Arxiv and your paper got featured on Hugging Face: https://huggingface.co/papers/2508.20615.
The paper page lets people discuss about your paper and lets them find artifacts related to it (your models, datasets or demo for instance). You can also claim the paper as yours which will show up on your public profile at HF, add Github and project page URLs.

I noticed that the EmoCAST model checkpoint is already available on the Hugging Face Hub at https://huggingface.co/Jmaomao/EmoCAST – that's fantastic! It would be even better to officially link it to your paper page on Hugging Face. This significantly improves its visibility and discoverability for other researchers and practitioners.

Additionally, your abstract mentions that you "construct an emotional talking head dataset". Would you be interested in hosting this new dataset on https://huggingface.co/datasets as well? We see it mentioned in the paper but couldn't find a direct download link for it.

Hosting your dataset on Hugging Face would enable people to easily load it using datasets.load_dataset():

from datasets import load_dataset

dataset = load_dataset("your-hf-org-or-username/your-dataset")

See here for a guide on uploading datasets: https://huggingface.co/docs/datasets/loading. The dataset viewer feature (https://huggingface.co/docs/hub/en/datasets-viewer) would also allow quick exploration of your data in the browser.

After uploading, we can link both the model and the dataset to your paper page (read here: https://huggingface.co/docs/hub/en/model-cards#linking-a-paper) so people can discover all your work in one place.

Let me know if you're interested or need any help regarding this!

Cheers,

Niels
ML Engineer @ HF 🤗