# SAM3RL: Memory Control via Reinforcement Learning for Visual Object Tracking

arXiv: https://arxiv.org/abs/2610.19931 · HTML: https://mlsanity.com/paper/2610.19931/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in RL with diffusion and flow models, one of 15 directions in Deep reinforcement learning methods.
- Builds on 25 of its 43 references that are in the map.
- First posted on arXiv 14 Oct 2026.

## Placed in

- Deep reinforcement learning methods / [RL with diffusion and flow models](https://mlsanity.com/deeprl/rl-with-diffusion-and-flow-models/)

## Builds on (most cited first, 25 of 25)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [SAM 2: Segment Anything in Images and Videos](https://mlsanity.com/paper/2408.00714/) (2024)
- [High-Dimensional Continuous Control Using Generalized Advantage Estimation](https://mlsanity.com/paper/1506.02438/) (2015)
- [DINOv3](https://mlsanity.com/paper/2508.10104/) (2025)
- [The 2017 DAVIS Challenge on Video Object Segmentation](https://mlsanity.com/paper/1704.00675/) (2017)
- [SAM 3: Segment Anything with Concepts](https://mlsanity.com/paper/2511.16719/) (2025)
- [GOT-10k: A Large High-Diversity Benchmark for Generic Object Tracking in the Wild](https://mlsanity.com/paper/1810.11981/) (2018)
- [Video Object Segmentation using Space-Time Memory Networks](https://mlsanity.com/paper/1904.00607/) (2019)
- [LaSOT: A High-quality Benchmark for Large-scale Single Object Tracking](https://mlsanity.com/paper/1809.07845/) (2018)
- [MixFormer: End-to-End Tracking with Iterative Mixed Attention](https://mlsanity.com/paper/2203.11082/) (2022)
- [YouTube-VOS: Sequence-to-Sequence Video Object Segmentation](https://mlsanity.com/paper/1809.00461/) (2018)
- [LaSOT: A High-quality Large-scale Single Object Tracking Benchmark](https://mlsanity.com/paper/2009.03465/) (2020)
- [High-Performance Long-Term Tracking with Meta-Updater](https://mlsanity.com/paper/2004.00305/) (2020)
- [SAM2Long: Enhancing SAM 2 for Long Video Segmentation with a Training-Free Memory Tree](https://mlsanity.com/paper/2410.16268/) (2024)
- [Learning the Model Update for Siamese Trackers](https://mlsanity.com/paper/1908.00855/) (2019)
- [MOSEv2: A More Challenging Dataset for Video Object Segmentation in Complex Scenes](https://mlsanity.com/paper/2508.05630/) (2025)
- [Deep Learning for Visual Tracking: A Comprehensive Survey](https://mlsanity.com/paper/1912.00535/) (2019)
- [Learning Dynamic Memory Networks for Object Tracking](https://mlsanity.com/paper/1803.07268/) (2018)
- [Visual Object Tracking with Discriminative Filters and Siamese Networks: A Survey and Outlook](https://mlsanity.com/paper/2112.02838/) (2021)
- [LVOS: A Benchmark for Large-scale Long-term Video Object Segmentation](https://mlsanity.com/paper/2404.19326/) (2024)
- [HiM2SAM: Enhancing SAM2 with Hierarchical Motion Estimation and Memory Optimization towards Long-term Tracking](https://mlsanity.com/paper/2507.07603/) (2025)
- [SAMITE: Position Prompted SAM2 with Calibrated Memory for Visual Object Tracking](https://mlsanity.com/paper/2507.21732/) (2025)
- [SAM2RL: Towards Reinforcement Learning Memory Control in Segment Anything Model 2](https://mlsanity.com/paper/2507.08548/) (2025)
- [Distractor-Aware Memory-Based Visual Object Tracking](https://mlsanity.com/paper/2509.13864/) (2025)
- [ACTrack: Adding Spatio-Temporal Condition for Visual Object Tracking](https://mlsanity.com/paper/2403.07914/) (2024)

## Abstract (arXiv)

Segment Anything Model 3 (SAM 3) is a foundation model for image and video segmentation that sets the state of the art in visual object tracking. Its tracker, inherited from SAM 2, stores information from past frames in a memory bank, managed by a simple First-In-First-Out rule, allowing temporal consistency in complex videos. Several recent methods improve SAM 2 with hand-crafted memory update rules, each targeting a specific failure mode, such as distractors, occlusions, or object motion. However, their gains transfer only partially to the stronger SAM 3 backbone. We instead present a more general, data-driven approach that frames memory control as a sequential decision-making task. SAM3RL is an extension of the SAM 3 tracker with a lightweight memory controller, trained by reinforcement learning, that decides which frames enter the memory bank in order to maximize tracking performance. SAM3RL sets a new state of the art on 10 of 12 tracking and video object segmentation benchmarks, outperforming all hand-crafted rules without any runtime overhead. The same approach applied to SAM 2 gives SAM2RL, which outperforms hand-crafted rules on 9 of the 12 benchmarks, showing the generality of our method across model generations.