Transfer Learning and Mixup for Fine-Grained Few-Shot Fungi Classification
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866908445886644224 |
|---|---|
| author | Tam, Jason Kahei Gustineli, Murilo Miyaguchi, Anthony |
| author_facet | Tam, Jason Kahei Gustineli, Murilo Miyaguchi, Anthony |
| contents | Accurate identification of fungi species presents a unique challenge in computer vision due to fine-grained inter-species variation and high intra-species variation. This paper presents our approach for the FungiCLEF 2025 competition, which focuses on few-shot fine-grained visual categorization (FGVC) using the FungiTastic Few-Shot dataset. Our team (DS@GT) experimented with multiple vision transformer models, data augmentation, weighted sampling, and incorporating textual information. We also explored generative AI models for zero-shot classification using structured prompting but found them to significantly underperform relative to vision-based models. Our final model outperformed both competition baselines and highlighted the effectiveness of domain specific pretraining and balanced sampling strategies. Our approach ranked 35/74 on the private test set in post-completion evaluation, this suggests additional work can be done on metadata selection and domain-adapted multi-modal learning. Our code is available at https://github.com/dsgt-arc/fungiclef-2025. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_08248 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Transfer Learning and Mixup for Fine-Grained Few-Shot Fungi Classification Tam, Jason Kahei Gustineli, Murilo Miyaguchi, Anthony Computer Vision and Pattern Recognition Information Retrieval Machine Learning Accurate identification of fungi species presents a unique challenge in computer vision due to fine-grained inter-species variation and high intra-species variation. This paper presents our approach for the FungiCLEF 2025 competition, which focuses on few-shot fine-grained visual categorization (FGVC) using the FungiTastic Few-Shot dataset. Our team (DS@GT) experimented with multiple vision transformer models, data augmentation, weighted sampling, and incorporating textual information. We also explored generative AI models for zero-shot classification using structured prompting but found them to significantly underperform relative to vision-based models. Our final model outperformed both competition baselines and highlighted the effectiveness of domain specific pretraining and balanced sampling strategies. Our approach ranked 35/74 on the private test set in post-completion evaluation, this suggests additional work can be done on metadata selection and domain-adapted multi-modal learning. Our code is available at https://github.com/dsgt-arc/fungiclef-2025. |
| title | Transfer Learning and Mixup for Fine-Grained Few-Shot Fungi Classification |
| topic | Computer Vision and Pattern Recognition Information Retrieval Machine Learning |
| url | https://arxiv.org/abs/2507.08248 |