InFeR: Informed Failure Resilience in Learned Visual Navigation Control
This work addresses the need for failure-resilient autonomous navigation in robotics by enabling IL policies to autonomously recover from OOD failures, though it is an incremental extension of existing techniques (VIB and Grad-CAM).
InFeR introduces a framework for imitation learning policies that detect, localize, and recover from out-of-distribution failures without requiring failure or recovery demonstrations, achieving robust long-range navigation in complex real-world environments.
While imitation learning (IL) has enabled successful visual navigation in many common environments, IL policies are prone to unpredictable failures under out-of-distribution (OOD) scenarios. This necessitates failure-resilient policies, which not only detect failures, but also recognise their sources and recover from them autonomously. We propose InFeR, a general framework for building IL policies with informed failure resilience without failure or recovery demonstrations. InFeR retrains an IL policy with a Variational Information Bottleneck (VIB) loss to structure its latent space for OOD failure detection. It applies a visual explainability technique, Grad-CAM, to localise an image region as the source of failure and inform a heuristic policy for recovery. All these are achieved without requiring additional training data. Real-world experiments show that InFeR enables informed failure recovery across two different policy architectures, yielding robust long-range navigation in complex environments.