Spatially Stratified Distillation for Heterogeneous Radar Place Recognition
For robotics and autonomous systems requiring all-weather place recognition across different radar hardware, this work improves cross-sensor matching performance.
The paper addresses heterogeneous radar place recognition, specifically matching sparse 4D automotive radar queries against dense spinning radar maps. The proposed spatially-stratified distillation (SSD) method achieves state-of-the-art results on the HeRCULES dataset, significantly outperforming prior methods on challenging dynamic sequences.
Scalable, all-weather place recognition increasingly relies on heterogeneous radar place recognition to bridge diverse hardware platforms. A notable application is matching queries from cost-effective 4D automotive radars against high-fidelity reference maps built by dense spinning radars. This process is fundamentally limited by the extreme sparsity (and narrow field-of-view) of the 4D sensor, which captures only a fraction of the structural density present in the spinning radar database. Prior efforts address this issue by unifying different radar signals. That is, projecting both signals into a common representational space. Yet, they suffer performance degradation in multi-session environments. In this paper, we propose spatially-stratified distillation (SSD); a strategy that replaces standard uniform distillation with an asymmetric spatial alignment derived directly from physical radar returns. In regions where both radars exhibit overlapping returns, SSD enforces strong feature alignment. Crucially, in sparse regions where the 4D student lacks returns but the teacher contains valid structure within the shared field of view, SSD applies heavily discounted distillation weights. Extensive evaluations of the recent HeRCULES dataset demonstrate that SSD significantly outperforms prior place recognition methods, achieving state-of-the-art results on its challenging dynamic sequences.