A Cross-Dataset Study for Text-Based 3D Human Motion Retrieval
Abstract
We provide results of our study on text-based 3D human motion retrieval and particularly focus on cross-dataset gen-eralization. Due to practical reasons such as dataset-specific human body representations, existing works typically benchmark by training and testing on partitions from the same dataset. Here, we employ a unified SMPL body format for all datasets, which allows us to perform training on one dataset, testing on the other, as well as training on a combination of datasets. Our results suggest that there exist dataset biases in standard text-motion benchmarks such as HumanML3D, KIT Motion-Language, and BABEL. We show that text augmentations help close the domain gap to some extent, but the gap remains. We further provide the first zero-shot action recognition results on BABEL, without using categorical action labels during training, opening up a new avenue for future research.
Cite
Text
Bensabath et al. "A Cross-Dataset Study for Text-Based 3D Human Motion Retrieval." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2024. doi:10.1109/CVPRW63382.2024.00199Markdown
[Bensabath et al. "A Cross-Dataset Study for Text-Based 3D Human Motion Retrieval." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2024.](https://mlanthology.org/cvprw/2024/bensabath2024cvprw-crossdataset/) doi:10.1109/CVPRW63382.2024.00199BibTeX
@inproceedings{bensabath2024cvprw-crossdataset,
title = {{A Cross-Dataset Study for Text-Based 3D Human Motion Retrieval}},
author = {Bensabath, Léore and Petrovich, Mathis and Varol, Gül},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
year = {2024},
pages = {1932-1940},
doi = {10.1109/CVPRW63382.2024.00199},
url = {https://mlanthology.org/cvprw/2024/bensabath2024cvprw-crossdataset/}
}