ACCOR: Attention-Enhanced Complex-Valued Contrastive Learning for Occluded Object Classification Using mmWave Radar IQ Signals

TL;DR

This paper introduces ACCOR, a novel attention-enhanced complex-valued contrastive learning method for mmWave radar that significantly improves occluded object classification accuracy across different frequencies.

Contribution

The paper presents ACCOR, a new complex-valued CNN with attention and hybrid loss, extending radar datasets and outperforming prior models in occluded object classification.

Findings

Achieves 96.60% accuracy at 64 GHz

Achieves 93.59% accuracy at 67 GHz

Outperforms prior radar-specific and image-based models

Abstract

Millimeter-wave (mmWave) radar provides robust sensing under adverse conditions and can penetrate thin materials for non-visual perception in industrial and robotic settings. Recent work with MIMO mmWave radar has demonstrated its ability to penetrate cardboard packaging for occluded object classification. However, existing models leave room for improvement and extensions across different sensing frequencies. Building on recent work with MIMO radar for occluded object classification, we propose ACCOR, an attention-enhanced complex-valued contrastive learning approach for radar, enabling robust occluded object classification. ACCOR processes complex-valued IQ radar signals via a complex-valued CNN backbone, a multi-head attention layer and a hybrid loss. The hybrid loss combines a weighted cross-entropy term with a supervised contrastive term. We extend an existing 64 GHz dataset with a new 67 GHz subset and evaluate performance across both bands. ACCOR achieves 96.60 % accuracy at 64 GHz and 93.59 % at 67 GHz on 10 objects, surpassing prior radar-specific and adapted image models. Results demonstrate the benefits of integrating complex-valued deep learning, attention, and contrastive learning for mmWave radar-based occluded object classification.