Do Object Channels Improve Robustness in Deep Reinforcement Learning?

Abstract:

Pixel-based reinforcement learning agents often exploit spurious visual correlations, leading to brittle policies that fail under minor visual perturbations. We systematically investigate spatial grounded semantic channel representations, often called Feature Maps, Planes, or Object Channels, as a representation design principle for reducing shortcut learning.
Object channels map detected entities into binary tensors aligned with the original coordinate frame, preserving compatibility with standard RL backbones without architectural modifications.
Specifically, through systematic evaluation in Atari environments under controlled perturbations, we demonstrate that such channel representations substantially improve zero-shot robustness to distribution shifts while maintaining competitive in-distribution performance.
We analyze the abstraction–fidelity trade-off and show that combining object channels with raw pixels improves robustness and sample efficiency compared to pure pixel-based approaches. The experimental results indicate that spatially grounded object-based encodings offer a practical mechanism for bridging pixel- and object-centric RL.

Submission Type: Regular submission (no more than 12 pages of main content)

Assigned Action Editor: ~Lijun_Wu1

Submission Number: 7847