Temporal Gradient Inversion for Private Trajectory Reconstruction in Embodied Reinforcement Learning
TRACE reconstructs private embodied observation-action trajectories from policy gradients, reaching 18.8 dB PSNR.
TRACE is an amortized temporal gradient-inversion attack that autoregressively reconstructs private observation-action trajectories from per-step policy gradients in embodied reinforcement learning. It uses cross-time correlation between successive gradients and closed-form action recovery from policy-head gradients when entropy regularization is small. On held-out scenes it reaches 18.8 dB PSNR and near-perfect action recovery at 3–4.5 ms per frame, beating learning-based and optimization baselines. The attack extends to recurrent, residual, and compact transformer victims, multimodal inputs, and larger discrete action spaces.
- TRACE inverts temporal policy gradients to recover private observation-action trajectories.
- Held-out embodied scenes reach 18.8 dB PSNR with near-perfect action recovery.
- Each frame reconstructs in 3 to 4.5 milliseconds, faster than optimization attacks.
- Authors argue temporal gradient streams need sequence-aware privacy defenses.
Full article179 words · extracted from arxiv.org · click to collapse
Distributed learning in embodied reinforcement-learning agents offers a degree of privacy by retaining raw sensor data on-device and transmitting only policy gradients to the server. Yet temporal structure can amplify this leakage beyond single-frame attacks. We introduce Temporal Reconstruction Attack on Consecutive Encodings (TRACE), an amortized temporal gradient-inversion attack that autoregressively reconstructs the sequence of private observation-action trajectories from per-step policy-learning gradients. The attack exploits two structural signals ignored by prior single-frame methods: (i) cross-time correlation between successive embodied gradients, which we formalize via a conditional mutual-information bound, and (ii) closed-form action recovery from policy-head gradient structure, which we prove exact when standard entropy regularization is sufficiently small. On held-out embodied scenes, TRACE reaches $18.8$ dB PSNR with near-perfect action recovery at $3$-$4.5$ ms per reconstructed frame, dominating the learning-based baseline across all reconstruction metrics and exceeding optimization attacks while running orders of magnitude faster. Further evaluation demonstrates TRACE's broader applicability across recurrent, residual, and compact transformer victim architectures, multi-modal inputs, and larger discrete action spaces. Defense experiments suggest that protecting temporal gradient streams may require sequence-aware privacy mechanisms.
Text extracted automatically; images, tables and formatting may be missing. Original: https://arxiv.org/abs/2609.30258