← Back to Benchmarks
simmediumnavigationmetric · varies
Nipping the Drift in the Bud: Retrospective Rectification for Robust Vision-Language Navigation
Description
Vision-Language Navigation (VLN) requires embodied agents to interpret natural language instructions and navigate through complex continuous 3D environments. However, the dominant imitation learning paradigm suffers from exposure bias, where minor deviations during inference lead to compounding errors. While DAgger-style approaches attempt to mitigate this by correcting error states, we identify a critical limitation: Instruction-State Misalignment. Forcing an agent to learn recovery actions fro