Efficient LiDAR Reflectance Compression via Scanning Serialization
This addresses the underexplored challenge of LiDAR reflectance compression for autonomous driving and robotics applications, offering a novel method with significant efficiency gains.
The paper tackles the problem of compressing LiDAR reflectance data by introducing SerLiC, a serialization-based neural compression framework that transforms 3D point clouds into 1D sequences for efficient modeling. The result is over 2x volume reduction, outperforming the state-of-the-art by up to 22% in compressed bits with only 2% of the parameters, and a lightweight version achieves >10 fps with 111K parameters.
Reflectance attributes in LiDAR point clouds provide essential information for downstream tasks but remain underexplored in neural compression methods. To address this, we introduce SerLiC, a serialization-based neural compression framework to fully exploit the intrinsic characteristics of LiDAR reflectance. SerLiC first transforms 3D LiDAR point clouds into 1D sequences via scan-order serialization, offering a device-centric perspective for reflectance analysis. Each point is then tokenized into a contextual representation comprising its sensor scanning index, radial distance, and prior reflectance, for effective dependencies exploration. For efficient sequential modeling, Mamba is incorporated with a dual parallelization scheme, enabling simultaneous autoregressive dependency capture and fast processing. Extensive experiments demonstrate that SerLiC attains over 2x volume reduction against the original reflectance data, outperforming the state-of-the-art method by up to 22% reduction of compressed bits while using only 2% of its parameters. Moreover, a lightweight version of SerLiC achieves > 10 fps (frames per second) with just 111K parameters, which is attractive for real-world applications.