Motion-Aware Reinforcement Learning For Object Localization

arXiv 2606.21764 ↗First posted 19 Jun 2026cs.CVcited by 0

On the map: Deep reinforcement learning methods → RL for robot locomotion and manipulation

Abstract (arXiv)

We present MARLNet (Motion-Aware Reinforcement Learning Network), a PPO-based bounding-box refinement agent that incorporates a constant-velocity motion prior into the observation state and an action smoothness penalty into the reward function. The agent operates on 268-dimensional observations encoding the current proposal, a kinematic prediction, the previous action, and a 256-dimensional EfficientNet-B0 crop feature, and learns a five-dimensional policy controlling coordinate adjustments and a binary termination trigger. Evaluated on Pascal VOC 2012 and VisDrone 2019, MARLNet trains stably across all regularization strengths tested and achieves consistent gains in detection success rate at IoU ≥ 0.5: up to +0.011 on VOC (λphys=0.10), where the motion prior prevents the overshooting that causes plain PPO to regress on this metric, and +0.007 on VisDrone (λphys=0.70), where unconstrained PPO achieves a larger gain (+0.025) owing to the weaker base detector. Through reward design ablations and training dynamics analysis, we identify a reward interference in which combining a constant-velocity deviation penalty with an absolute IoU term causes trigger collapse, and show that replacing it with the action smoothness penalty resolves this failure. We further characterize a representational ceiling facing crop-feature refinement agents that share a backbone with their base detector, confirmed through a global-plus-local observation ablation. Project page: https://prithviraj97.github.io/marl-net

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in RL for robot locomotion and manipulation, one of 15 directions in Deep reinforcement learning methods. Builds on 19 of its 32 references that are in the map. First posted on arXiv 19 Jun 2026.

Builds on

19 of 32 references.

References · 19

2015Deep Residual Learning for Image Recognitioncited by 52,096
2014Microsoft COCO: Common Objects in Contextcited by 17,608
2017Proximal Policy Optimization Algorithmscited by 15,215
2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,639
2015High-Dimensional Continuous Control Using Generalized Advantage Estimationcited by 2,026
2016Simple Online and Realtime Trackingcited by 868
2020Detection and Tracking Meet Drones Challengecited by 263
2018Visual Foresight: Model-Based Deep Reinforcement Learning for Vision-Based Robotic Controlcited by 230
2020Kinematic 3D Object Detection in Monocular Videocited by 116
2020Alpha-Refine: Boosting Tracking Performance by Precise Bounding Box Estimationcited by 81
2020Regularizing Action Policies for Smooth Control with Reinforcement Learningcited by 59
2015Active Object Localization with Deep Reinforcement Learningcited by 42

7 more

2023MotionTrack: Learning Motion Predictor for Multiple Object Trackingcited by 20
2024Gradient-based Regularization for Action Smoothness in Robotic Control with Reinforcement Learningcited by 10
2024Visible and Clear: Finding Tiny Objects in Difference Mapcited by 9
2024Beyond Kalman Filters: Deep Learning-Based Filters for Improved Object Trackingcited by 8
2023Traffic Video Object Detection using Motion Priorcited by 4
2018Multi-Stage Reinforcement Learning For Object Detectioncited by 2
2023Enhancing Your Trained DETRs with Box Refinementcited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way