Enhancing Remote Sensing Vision-Language Models for Zero-Shot Scene Classification
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916554642292736 |
|---|---|
| author | Khoury, Karim El Zanella, Maxime Gérin, Benoît Godelaine, Tiffanie Macq, Benoît Mahmoudi, Saïd De Vleeschouwer, Christophe Ayed, Ismail Ben |
| author_facet | Khoury, Karim El Zanella, Maxime Gérin, Benoît Godelaine, Tiffanie Macq, Benoît Mahmoudi, Saïd De Vleeschouwer, Christophe Ayed, Ismail Ben |
| contents | Vision-Language Models for remote sensing have shown promising uses thanks to their extensive pretraining. However, their conventional usage in zero-shot scene classification methods still involves dividing large images into patches and making independent predictions, i.e., inductive inference, thereby limiting their effectiveness by ignoring valuable contextual information. Our approach tackles this issue by utilizing initial predictions based on text prompting and patch affinity relationships from the image encoder to enhance zero-shot capabilities through transductive inference, all without the need for supervision and at a minor computational cost. Experiments on 10 remote sensing datasets with state-of-the-art Vision-Language Models demonstrate significant accuracy improvements over inductive zero-shot classification. Our source code is publicly available on Github: https://github.com/elkhouryk/RS-TransCLIP |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_00698 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Enhancing Remote Sensing Vision-Language Models for Zero-Shot Scene Classification Khoury, Karim El Zanella, Maxime Gérin, Benoît Godelaine, Tiffanie Macq, Benoît Mahmoudi, Saïd De Vleeschouwer, Christophe Ayed, Ismail Ben Computer Vision and Pattern Recognition Vision-Language Models for remote sensing have shown promising uses thanks to their extensive pretraining. However, their conventional usage in zero-shot scene classification methods still involves dividing large images into patches and making independent predictions, i.e., inductive inference, thereby limiting their effectiveness by ignoring valuable contextual information. Our approach tackles this issue by utilizing initial predictions based on text prompting and patch affinity relationships from the image encoder to enhance zero-shot capabilities through transductive inference, all without the need for supervision and at a minor computational cost. Experiments on 10 remote sensing datasets with state-of-the-art Vision-Language Models demonstrate significant accuracy improvements over inductive zero-shot classification. Our source code is publicly available on Github: https://github.com/elkhouryk/RS-TransCLIP |
| title | Enhancing Remote Sensing Vision-Language Models for Zero-Shot Scene Classification |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2409.00698 |