Residual Vector Quantization For Communication-Efficient Multi-Agent Perception
This addresses the scalability problem for autonomous vehicles and robots in multi-agent systems, enabling efficient and accurate collaborative perception with a step toward practical deployment, though it is incremental as it builds on existing compression and quantization techniques.
The paper tackles the communication bandwidth limitation in multi-agent collaborative perception by introducing ReVQom, a learned feature codec that compresses intermediate features from 8192 bpp to 6-30 bpp, achieving up to 1365x compression with minimal accuracy loss on the DAIR-V2X dataset.
Multi-agent collaborative perception (CP) improves scene understanding by sharing information across connected agents such as autonomous vehicles, unmanned aerial vehicles, and robots. Communication bandwidth, however, constrains scalability. We present ReVQom, a learned feature codec that preserves spatial identity while compressing intermediate features. ReVQom is an end-to-end method that compresses feature dimensions via a simple bottleneck network followed by multi-stage residual vector quantization (RVQ). This allows only per-pixel code indices to be transmitted, reducing payloads from 8192 bits per pixel (bpp) of uncompressed 32-bit float features to 6-30 bpp per agent with minimal accuracy loss. On DAIR-V2X real-world CP dataset, ReVQom achieves 273x compression at 30 bpp to 1365x compression at 6 bpp. At 18 bpp (455x), ReVQom matches or outperforms raw-feature CP, and at 6-12 bpp it enables ultra-low-bandwidth operation with graceful degradation. ReVQom allows efficient and accurate multi-agent collaborative perception with a step toward practical V2X deployment.