Competing for pixels: a self-play algorithm for weakly-supervised segmentation
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913364134854656 |
|---|---|
| author | Saeed, Shaheer U. Huang, Shiqi Ramalhinho, João Gayo, Iani J. M. B. Montaña-Brown, Nina Bonmati, Ester Pereira, Stephen P. Davidson, Brian Barratt, Dean C. Clarkson, Matthew J. Hu, Yipeng |
| author_facet | Saeed, Shaheer U. Huang, Shiqi Ramalhinho, João Gayo, Iani J. M. B. Montaña-Brown, Nina Bonmati, Ester Pereira, Stephen P. Davidson, Brian Barratt, Dean C. Clarkson, Matthew J. Hu, Yipeng |
| contents | Weakly-supervised segmentation (WSS) methods, reliant on image-level labels indicating object presence, lack explicit correspondence between labels and regions of interest (ROIs), posing a significant challenge. Despite this, WSS methods have attracted attention due to their much lower annotation costs compared to fully-supervised segmentation. Leveraging reinforcement learning (RL) self-play, we propose a novel WSS method that gamifies image segmentation of a ROI. We formulate segmentation as a competition between two agents that compete to select ROI-containing patches until exhaustion of all such patches. The score at each time-step, used to compute the reward for agent training, represents likelihood of object presence within the selection, determined by an object presence detector pre-trained using only image-level binary classification labels of object presence. Additionally, we propose a game termination condition that can be called by either side upon exhaustion of all ROI-containing patches, followed by the selection of a final patch from each. Upon termination, the agent is incentivised if ROI-containing patches are exhausted or disincentivised if an ROI-containing patch is found by the competitor. This competitive setup ensures minimisation of over- or under-segmentation, a common problem with WSS methods. Extensive experimentation across four datasets demonstrates significant performance improvements over recent state-of-the-art methods. Code: https://github.com/s-sd/spurl/tree/main/wss |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2405_16628 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Competing for pixels: a self-play algorithm for weakly-supervised segmentation Saeed, Shaheer U. Huang, Shiqi Ramalhinho, João Gayo, Iani J. M. B. Montaña-Brown, Nina Bonmati, Ester Pereira, Stephen P. Davidson, Brian Barratt, Dean C. Clarkson, Matthew J. Hu, Yipeng Computer Vision and Pattern Recognition Machine Learning Weakly-supervised segmentation (WSS) methods, reliant on image-level labels indicating object presence, lack explicit correspondence between labels and regions of interest (ROIs), posing a significant challenge. Despite this, WSS methods have attracted attention due to their much lower annotation costs compared to fully-supervised segmentation. Leveraging reinforcement learning (RL) self-play, we propose a novel WSS method that gamifies image segmentation of a ROI. We formulate segmentation as a competition between two agents that compete to select ROI-containing patches until exhaustion of all such patches. The score at each time-step, used to compute the reward for agent training, represents likelihood of object presence within the selection, determined by an object presence detector pre-trained using only image-level binary classification labels of object presence. Additionally, we propose a game termination condition that can be called by either side upon exhaustion of all ROI-containing patches, followed by the selection of a final patch from each. Upon termination, the agent is incentivised if ROI-containing patches are exhausted or disincentivised if an ROI-containing patch is found by the competitor. This competitive setup ensures minimisation of over- or under-segmentation, a common problem with WSS methods. Extensive experimentation across four datasets demonstrates significant performance improvements over recent state-of-the-art methods. Code: https://github.com/s-sd/spurl/tree/main/wss |
| title | Competing for pixels: a self-play algorithm for weakly-supervised segmentation |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2405.16628 |