Fine-Grained Classification: Connecting Metadata via Cross-Contrastive Pre-Training
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_ | 1866911218024841216 |
|---|---|
| author | Mamtani, Sumit Thesia, Yash |
| author_facet | Mamtani, Sumit Thesia, Yash |
| contents | Fine-grained visual classification aims to recognize objects belonging to many subordinate categories of a supercategory, where appearance alone often fails to distinguish highly similar classes. We propose a unified framework that integrates image, text, and metadata via cross-contrastive pre-training. We first align the three modality encoders in a shared embedding space and then fine-tune the image and metadata encoders for classification. On NABirds, our approach improves over the baseline by 7.83% and achieves 84.44% top-1 accuracy, outperforming strong multimodal methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_20322 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Fine-Grained Classification: Connecting Metadata via Cross-Contrastive Pre-Training Mamtani, Sumit Thesia, Yash Computer Vision and Pattern Recognition Machine Learning Fine-grained visual classification aims to recognize objects belonging to many subordinate categories of a supercategory, where appearance alone often fails to distinguish highly similar classes. We propose a unified framework that integrates image, text, and metadata via cross-contrastive pre-training. We first align the three modality encoders in a shared embedding space and then fine-tune the image and metadata encoders for classification. On NABirds, our approach improves over the baseline by 7.83% and achieves 84.44% top-1 accuracy, outperforming strong multimodal methods. |
| title | Fine-Grained Classification: Connecting Metadata via Cross-Contrastive Pre-Training |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2504.20322 |