SOURCE-LINKED INTELLIGENCE

Visual Navigation Transformer with Pose Attention

arXiv · AI, language, vision and robotics · article · Sep 18, 2026 · UTC

Learned navigation policies typically consume observations as a temporally ordered history, with positional encodings tying each observation to when it was seen, making it difficult to reuse experience from earlier traversals of an environment. Systems that do reuse such experience usually construct an explicit representation, such as a map or a topological graph, and plan on it. We propose VNT-PA (Visual Navigation Transformer with Pose Attention), a transformer planner whose context is a set of depth keyframes indexed by camera pose. With camera poses as positional encoding, attention depend

Read original source ↗ Open in workspace

recordType

paper

region

Global