Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2512.05996 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915657437675520 |
|---|---|
| author | Liu, Yi Song, Jingyu Kallakuri, Vedanth Skinner, Katherine A. |
| author_facet | Liu, Yi Song, Jingyu Kallakuri, Vedanth Skinner, Katherine A. |
| contents | Analyzing underwater fish imagery is critical for ecological monitoring but remains difficult due to visual degradation and costly annotations. We introduce FishDetector-R1, a unified MLLM-based framework for fish detection, segmentation, and counting under weak supervision. On the DeepFish dataset, our framework achieves substantial gains over baselines, improving AP by 20% and mIoU by 10%, while reducing MAE by 30% and GAME by 35%. These improvements stem from two key components: a novel detect-to-count prompt that enforces spatially consistent detections and counts, and Reinforcement Learning from Verifiable Reward (RLVR) with a complementary scalable paradigm leveraging sparse point labels. Ablation studies further validate the effectiveness of this reward design. Moreover, the improvement generalizes well to other underwater datasets, confirming strong cross-domain robustness. Overall, FishDetector-R1 provides a reliable and scalable solution for accurate marine visual understanding via weak supervision. The project page for FishDetector-R1 is https://umfieldrobotics.github.io/FishDetector-R1. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_05996 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | FishDetector-R1: Unified MLLM-Based Framework with Reinforcement Fine-Tuning for Weakly Supervised Fish Detection, Segmentation, and Counting Liu, Yi Song, Jingyu Kallakuri, Vedanth Skinner, Katherine A. Computer Vision and Pattern Recognition Computers and Society Robotics Image and Video Processing Analyzing underwater fish imagery is critical for ecological monitoring but remains difficult due to visual degradation and costly annotations. We introduce FishDetector-R1, a unified MLLM-based framework for fish detection, segmentation, and counting under weak supervision. On the DeepFish dataset, our framework achieves substantial gains over baselines, improving AP by 20% and mIoU by 10%, while reducing MAE by 30% and GAME by 35%. These improvements stem from two key components: a novel detect-to-count prompt that enforces spatially consistent detections and counts, and Reinforcement Learning from Verifiable Reward (RLVR) with a complementary scalable paradigm leveraging sparse point labels. Ablation studies further validate the effectiveness of this reward design. Moreover, the improvement generalizes well to other underwater datasets, confirming strong cross-domain robustness. Overall, FishDetector-R1 provides a reliable and scalable solution for accurate marine visual understanding via weak supervision. The project page for FishDetector-R1 is https://umfieldrobotics.github.io/FishDetector-R1. |
| title | FishDetector-R1: Unified MLLM-Based Framework with Reinforcement Fine-Tuning for Weakly Supervised Fish Detection, Segmentation, and Counting |
| topic | Computer Vision and Pattern Recognition Computers and Society Robotics Image and Video Processing |
| url | https://arxiv.org/abs/2512.05996 |