SALSA: Swift Adaptive Lightweight Self-Attention for Enhanced LiDAR Place Recognition

TL;DR

SALSA is a new lightweight LiDAR place recognition framework that uses adaptive self-attention and a Sphereformer backbone to improve accuracy and efficiency in large-scale mapping and localization tasks.

Contribution

It introduces a novel Sphereformer backbone with radial window attention and an adaptive self-attention layer for enhanced scene representation in LiDAR place recognition.

Findings

Outperforms existing methods in retrieval accuracy

Achieves better metric localization results

Operates in real-time with high efficiency

Abstract

Large-scale LiDAR mappings and localization leverage place recognition techniques to mitigate odometry drifts, ensuring accurate mapping. These techniques utilize scene representations from LiDAR point clouds to identify previously visited sites within a database. Local descriptors, assigned to each point within a point cloud, are aggregated to form a scene representation for the point cloud. These descriptors are also used to re-rank the retrieved point clouds based on geometric fitness scores. We propose SALSA, a novel, lightweight, and efficient framework for LiDAR place recognition. It consists of a Sphereformer backbone that uses radial window attention to enable information aggregation for sparse distant points, an adaptive self-attention layer to pool local descriptors into tokens, and a multi-layer-perceptron Mixer layer for aggregating the tokens to generate a scene descriptor. The proposed framework outperforms existing methods on various LiDAR place recognition datasets in terms of both retrieval and metric localization while operating in real-time.