Temporal Residual Bottleneck for Robust Asynchronous Collaborative Perception

Key points

• Most latency-robust methods compensate delayed collaborator features through flow-guided alignment or direct feature transport.1
• In this work, we formulate asynchronous collaborative perception as temporal residual prediction.1
• Our Temporal Residual Bottleneck keeps a deterministic pose-warped collaborator feature as a conservative anchor and uses a $Δt$-conditioned xLSTM to extract residual temporal evidence from the available history.1
• A detector-facing residual bottleneck then applies only gated, regularized corrections before ego-side fusion, reducing the risk of overwriting reliable static structure when temporal correspondence is uncertain.1

Sources (1)

• [1]Temporal Residual Bottleneck for Robust Asynchronous Collaborative Perception
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 7, 01:49 PM

  Collaborative perception extends the sensing range of autonomous vehicles, but its performance degrades when shared features arrive stale or incomplete.

  Most latency-robust methods compensate delayed collaborator features through flow-guided alignment or direct feature transport.

Extractive summary: sentences quoted from the sources.