10 papers
Federated Subspace Guided Vision-Language-Action Policy Distillation for Non-IID Multi-Robot Manipulation
Biprodip Pal, Kaushik Roy, Yanming Zhu +3
Federated learning offers a natural way for multiple robots to jointly improve manipulation policies without requiring centralized access to training demonstrations. However, non-I…
Geometry-Conditioned Visual Place Recognition in Natural Environments
Walter Nedov, Saimunur Rahman, Kavindie Katuwandeniya +3
Visual Place Recognition (VPR) in natural environments remains challenging due to repetitive vegetation, sparse distinctive landmarks, and substantial appearance and viewpoint vari…
It's Not Just More Demos: Counterfactual Action Sensitivity Coverage for Data-Efficient Robust Robot Imitation
Giovanni D'urso, Kaushik Roy, Nicholas Lawrance +1
Visuomotor imitation learning has demonstrated success for manipulation tasks. However, the trained policies remain brittle to visual `nuisances', with even minor task-preserving v…
TOLiD: Bridging the Architecture Gap in Vision Foundation Model to LiDAR Pretraining via Token Lifting for Distillation
Sutharsan Mahendran, Darshana Priyasad, Kaushik Roy +4
Cross-modal distillation from Vision Foundation Models (VFMs) to LiDAR backbones has recently emerged as a self-supervised pretraining strategy that reduces reliance on dense point…
Visual Place Recognition in Forests with Depth-Aware Distillation
Walter Nedov, Saimunur Rahman, Kavindie Katuwandeniya +3
Visual place recognition in natural forest environments remains challenging due to repetitive vegetation, weak structural cues, and significant appearance variation across traversa…
SPREAD: Subspace Representation Distillation for Lifelong Imitation Learning
Kaushik Roy, Giovanni D'urso, Nicholas Lawrance +2
A key challenge in lifelong imitation learning (LIL) is enabling agents to acquire new skills from expert demonstrations while retaining prior knowledge. This requires preserving t…
WildCross: A Cross-Modal Large Scale Benchmark for Place Recognition and Metric Depth Estimation in Natural Environments
Joshua Knights, Joseph Reid, Kaushik Roy +3
Recent years have seen a significant increase in demand for robotic solutions in unstructured natural environments, alongside growing interest in bridging 2D and 3D scene understan…
LANCE: Low Rank Activation Compression for Efficient On-Device Continual Learning
Marco Paul E. Apolinario, Kaushik Roy
On-device learning is essential for personalization, privacy, and long-term adaptation in resource-constrained environments. Achieving this requires efficient learning, both fine-t…
CODE-CL: Conceptor-Based Gradient Projection for Deep Continual Learning
Marco Paul E. Apolinario, Sakshi Choudhary, Kaushik Roy
Continual learning (CL) - the ability to progressively acquire and integrate new concepts - is essential to intelligent systems to adapt to dynamic environments. However, deep neur…
M2Distill: Multi-Modal Distillation for Lifelong Imitation Learning
Kaushik Roy, Akila Dissanayake, Brendan Tidd +1
Lifelong imitation learning for manipulation tasks poses significant challenges due to distribution shifts that occur in incremental learning steps. Existing methods often focus on…