S^4M: Boosting Semi-Supervised Instance Segmentation with SAM
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_ | 1866908305549426688 |
|---|---|
| author | Yoon, Heeji Shin, Heeseong Hong, Eunbeen Choi, Hyunwook Cho, Hansang Jeong, Daun Kim, Seungryong |
| author_facet | Yoon, Heeji Shin, Heeseong Hong, Eunbeen Choi, Hyunwook Cho, Hansang Jeong, Daun Kim, Seungryong |
| contents | Semi-supervised instance segmentation poses challenges due to limited labeled data, causing difficulties in accurately localizing distinct object instances. Current teacher-student frameworks still suffer from performance constraints due to unreliable pseudo-label quality stemming from limited labeled data. While the Segment Anything Model (SAM) offers robust segmentation capabilities at various granularities, directly applying SAM to this task introduces challenges such as class-agnostic predictions and potential over-segmentation. To address these complexities, we carefully integrate SAM into the semi-supervised instance segmentation framework, developing a novel distillation method that effectively captures the precise localization capabilities of SAM without compromising semantic recognition. Furthermore, we incorporate pseudo-label refinement as well as a specialized data augmentation with the refined pseudo-labels, resulting in superior performance. We establish state-of-the-art performance, and provide comprehensive experiments and ablation studies to validate the effectiveness of our proposed approach. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_05301 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | S^4M: Boosting Semi-Supervised Instance Segmentation with SAM Yoon, Heeji Shin, Heeseong Hong, Eunbeen Choi, Hyunwook Cho, Hansang Jeong, Daun Kim, Seungryong Computer Vision and Pattern Recognition Semi-supervised instance segmentation poses challenges due to limited labeled data, causing difficulties in accurately localizing distinct object instances. Current teacher-student frameworks still suffer from performance constraints due to unreliable pseudo-label quality stemming from limited labeled data. While the Segment Anything Model (SAM) offers robust segmentation capabilities at various granularities, directly applying SAM to this task introduces challenges such as class-agnostic predictions and potential over-segmentation. To address these complexities, we carefully integrate SAM into the semi-supervised instance segmentation framework, developing a novel distillation method that effectively captures the precise localization capabilities of SAM without compromising semantic recognition. Furthermore, we incorporate pseudo-label refinement as well as a specialized data augmentation with the refined pseudo-labels, resulting in superior performance. We establish state-of-the-art performance, and provide comprehensive experiments and ablation studies to validate the effectiveness of our proposed approach. |
| title | S^4M: Boosting Semi-Supervised Instance Segmentation with SAM |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2504.05301 |