HUWSOD: Holistic Self-training for Unified Weakly Supervised Object Detection
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_ | 1866914850690564096 |
|---|---|
| author | Cao, Liujuan Lin, Jianghang Hong, Zebo Shen, Yunhang Lin, Shaohui Chen, Chao Ji, Rongrong |
| author_facet | Cao, Liujuan Lin, Jianghang Hong, Zebo Shen, Yunhang Lin, Shaohui Chen, Chao Ji, Rongrong |
| contents | Most WSOD methods rely on traditional object proposals to generate candidate regions and are confronted with unstable training, which easily gets stuck in a poor local optimum. In this paper, we introduce a unified, high-capacity weakly supervised object detection (WSOD) network called HUWSOD, which utilizes a comprehensive self-training framework without needing external modules or additional supervision. HUWSOD innovatively incorporates a self-supervised proposal generator and an autoencoder proposal generator with a multi-rate resampling pyramid to replace traditional object proposals, enabling end-to-end WSOD training and inference. Additionally, we implement a holistic self-training scheme that refines detection scores and coordinates through step-wise entropy minimization and consistency-constraint regularization, ensuring consistent predictions across stochastic augmentations of the same image. Extensive experiments on PASCAL VOC and MS COCO demonstrate that HUWSOD competes with state-of-the-art WSOD methods, eliminating the need for offline proposals and additional data. The peak performance of HUWSOD approaches that of fully-supervised Faster R-CNN. Our findings also indicate that randomly initialized boxes, although significantly different from well-designed offline object proposals, are effective for WSOD training. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_19394 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | HUWSOD: Holistic Self-training for Unified Weakly Supervised Object Detection Cao, Liujuan Lin, Jianghang Hong, Zebo Shen, Yunhang Lin, Shaohui Chen, Chao Ji, Rongrong Computer Vision and Pattern Recognition Most WSOD methods rely on traditional object proposals to generate candidate regions and are confronted with unstable training, which easily gets stuck in a poor local optimum. In this paper, we introduce a unified, high-capacity weakly supervised object detection (WSOD) network called HUWSOD, which utilizes a comprehensive self-training framework without needing external modules or additional supervision. HUWSOD innovatively incorporates a self-supervised proposal generator and an autoencoder proposal generator with a multi-rate resampling pyramid to replace traditional object proposals, enabling end-to-end WSOD training and inference. Additionally, we implement a holistic self-training scheme that refines detection scores and coordinates through step-wise entropy minimization and consistency-constraint regularization, ensuring consistent predictions across stochastic augmentations of the same image. Extensive experiments on PASCAL VOC and MS COCO demonstrate that HUWSOD competes with state-of-the-art WSOD methods, eliminating the need for offline proposals and additional data. The peak performance of HUWSOD approaches that of fully-supervised Faster R-CNN. Our findings also indicate that randomly initialized boxes, although significantly different from well-designed offline object proposals, are effective for WSOD training. |
| title | HUWSOD: Holistic Self-training for Unified Weakly Supervised Object Detection |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2406.19394 |