AdaRadar: Rate Adaptive Spectral Compression for Radar-based Perception
2026-03-18 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors address the problem of sending large amounts of radar data in self-driving cars through limited communication links. They created a smart compression method that adjusts how much data to compress based on detection confidence, using an indirect way to estimate gradients without sending large extra data. They also use a mathematical transform to focus on important radar signal parts and reduce data size a lot while keeping performance nearly the same. Their method was tested successfully on multiple radar datasets.
radar data compressionautonomous drivingadaptive compressiongradient descentzeroth-order gradient approximationdiscrete cosine transformquantizationdata pruningrange-Dopplerfeature maps
Authors
Jinho Park, Se Young Chun, Mingoo Seok
Abstract
Radar is a critical perception modality in autonomous driving systems due to its all-weather characteristics and ability to measure range and Doppler velocity. However, the sheer volume of high-dimensional raw radar data saturates the communication link to the computing engine (e.g., an NPU), which is often a low-bandwidth interface with data rate provisioned only for a few low-resolution range-Doppler frames. A generalized codec for utilizing high-dimensional radar data is notably absent, while existing image-domain approaches are unsuitable, as they typically operate at fixed compression ratios and fail to adapt to varying or adversarial conditions. In light of this, we propose radar data compression with adaptive feedback. It dynamically adjusts the compression ratio by performing gradient descent from the proxy gradient of detection confidence with respect to the compression rate. We employ a zeroth-order gradient approximation as it enables gradient computation even with non-differentiable core operations--pruning and quantization. This also avoids transmitting the gradient tensors over the band-limited link, which, if estimated, would be as large as the original radar data. In addition, we have found that radar feature maps are heavily concentrated on a few frequency components. Thus, we apply the discrete cosine transform to the radar data cubes and selectively prune out the coefficients effectively. We preserve the dynamic range of each radar patch through scaled quantization. Combining those techniques, our proposed online adaptive compression scheme achieves over 100x feature size reduction at minimal performance drop (~1%p). We validate our results on the RADIal, CARRADA, and Radatron datasets.