Towards Training-Free Open-World Classification with 3D Generative Models
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866913796796186624 |
|---|---|
| author | Xia, Xinzhe Zhao, Weiguang Yan, Yuyao Yang, Guanyu Zhang, Rui Huang, Kaizhu Yang, Xi |
| author_facet | Xia, Xinzhe Zhao, Weiguang Yan, Yuyao Yang, Guanyu Zhang, Rui Huang, Kaizhu Yang, Xi |
| contents | 3D open-world classification is a challenging yet essential task in dynamic and unstructured real-world scenarios, requiring both open-category and open-pose recognition. To address these challenges, recent wisdom often takes sophisticated 2D pre-trained models to provide enriched and stable representations. However, these methods largely rely on how 3D objects can be projected into 2D space, which is unfortunately not well solved, and thus significantly limits their performance. Unlike these present efforts, in this paper we make a pioneering exploration of 3D generative models for 3D open-world classification. Drawing on abundant prior knowledge from 3D generative models, we additionally craft a rotation-invariant feature extractor. This innovative synergy endows our pipeline with the advantages of being training-free, open-category, and pose-invariant, thus well suited to 3D open-world classification. Extensive experiments on benchmark datasets demonstrate the potential of generative models in 3D open-world classification, achieving state-of-the-art performance on ModelNet10 and McGill with 32.0% and 8.7% overall accuracy improvement, respectively. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_17547 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Towards Training-Free Open-World Classification with 3D Generative Models Xia, Xinzhe Zhao, Weiguang Yan, Yuyao Yang, Guanyu Zhang, Rui Huang, Kaizhu Yang, Xi Computer Vision and Pattern Recognition 3D open-world classification is a challenging yet essential task in dynamic and unstructured real-world scenarios, requiring both open-category and open-pose recognition. To address these challenges, recent wisdom often takes sophisticated 2D pre-trained models to provide enriched and stable representations. However, these methods largely rely on how 3D objects can be projected into 2D space, which is unfortunately not well solved, and thus significantly limits their performance. Unlike these present efforts, in this paper we make a pioneering exploration of 3D generative models for 3D open-world classification. Drawing on abundant prior knowledge from 3D generative models, we additionally craft a rotation-invariant feature extractor. This innovative synergy endows our pipeline with the advantages of being training-free, open-category, and pose-invariant, thus well suited to 3D open-world classification. Extensive experiments on benchmark datasets demonstrate the potential of generative models in 3D open-world classification, achieving state-of-the-art performance on ModelNet10 and McGill with 32.0% and 8.7% overall accuracy improvement, respectively. |
| title | Towards Training-Free Open-World Classification with 3D Generative Models |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2501.17547 |