Laser2Vec: Similarity-based Retrieval for Robotic Perception Data

TL;DR

Laser2Vec introduces a similarity-based retrieval system for robotic perception data, enabling efficient search of 2D LiDAR scans through learned compressed representations, improving analysis of large robotic datasets.

Contribution

The paper presents a novel system combining convolutional variational autoencoders and a lightweight distance approximation network for fast, accurate retrieval of similar robotic perception scans.

Findings

System accurately identifies similar scans in real-world deployments

Efficient retrieval outperforms linear scan methods

Robust across diverse indoor environments

Abstract

As mobile robot capabilities improve and deployment times increase, tools to analyze the growing volume of data are becoming necessary. Current state-of-the-art logging, playback, and exploration systems are insufficient for practitioners seeking to discover systemic points of failure in robotic systems. This paper presents a suite of algorithms for similarity-based queries of robotic perception data and implements a system for storing 2D LiDAR data from many deployments cheaply and evaluating top-k queries for complete or partial scans efficiently. We generate compressed representations of laser scans via a convolutional variational autoencoder and store them in a database, where a light-weight dense network for distance function approximation is run at query time. Our query evaluator leverages the local continuity of the embedding space to generate evaluation orders that, in expectation, dominate full linear scans of the database. The accuracy, robustness, scalability, and efficiency of our system is tested on real-world data gathered from dozens of deployments and synthetic data generated by corrupting real data. We find our system accurately and efficiently identifies similar scans across a number of episodes where the robot encountered the same location, or similar indoor structures or objects.