SeqNet: Learning Descriptors for Sequence-Based Hierarchical Place Recognition

摘要

Learns sequence descriptors for hierarchical visual place recognition combining single-image and sequence-level matching.