MMS-VPR: A Fine-Grained Multimodal Street-Level Visual Place Recognition Dataset and Evaluation Benchmark for Dense Pedestrian Environments
A unified benchmarking platform consolidating 22 baselines spanning shallow ML, CNN, Transformer, graph neural network (GNN), and multimodal architectures under a standardized, reproducible pipeline for pedestrian-level place recognition in dense urban environments is released.
Yi-Wei Ou, Xiao-Bin Ren, Rong Sun et al.
· 5 citations