Open-RadVLAD: Fast and Robust Radar Place Recognition

Radar place recognition often involves encoding a live scan as a vector and matching this vector to a database in order to recognise that the vehicle is in a location that it has visited before. Radar is inherently robust to lighting or weather conditions, but place recognition with this sensor is still affected by: (1) viewpoint variation, i.e. translation and rotation, (2) sensor artefacts or “noises”. For 360° scanning radar, rotation is readily dealt with by in some way aggregating across azimuths. Also, we argue in this work that it is more critical to deal with the richness and informativeness of representation than it is to deal with translational invariance – particularly in urban driving where vehicles predominantly follow the same lane when repeating a route. In our method, for computational efficiency, we use only the polar representation. For partial translation invariance, we use only a one-dimensional Fourier Transform along radial returns. As the original radar signal is in the form of received power in discretised range bins, we also show experimentally that taking a radial Fourier Transform in this way and matching based on spatial frequencies present in the power signal leads to better performance – leading to a 7% to 8% improvement in localisation success (Section IV and Table II). We also achieve rotational invariance and a very discriminative descriptor space by building a vector of locally aggregated descriptors (VLAD). Our method is more comprehensively tested than all prior radar place recognition work – over an exhaustive combination of all 870 pairs of trajectories from 30 Oxford Radar RobotCar Dataset sequences (each ≈10 km), with a frequency-modulated continuous-wave (FMCW) radar. Code and detailed results are provided at github.com/mttgdd/open-radvlad, as an open implementation and benchmark for future work in this area. We achieve a mean of 89.35% and median of 91.52% in Recall@1, outstripping the mean of 68.56% and median of 69.55% for the only other open implementation, RaPlace, and at a fraction of its computational cost (relying on fewer integral transforms e.g. Radon, Fourier, and inverse Fourier).