DeepPCC: Learned Lossy Point Cloud Compression

Junzhe Zhang,Gexin Liu,Junteng Zhang,Dandan Ding,Zhan Ma

DOI: https://doi.org/10.1109/tetci.2024.3467192

2024-01-01

IEEE Transactions on Emerging Topics in Computational Intelligence

Abstract:We propose DeepPCC, an end-to-end learning-based approach for the lossy compression of large-scale object point clouds. For both geometry and attribute components, we introduce the Multiscale Neighborhood Information Aggregation (NIA) mechanism, which applies resolution downscaling progressively ( i.e. , dyadic downsampling of geometry and average pooling of attribute) and combines sparse convolution and local self-attention at each resolution scale for effective feature representation. Under a simple autoencoder structure, scale-wise NIA blocks are stacked as the analysis and synthesis transform in the encoder-decoder pair to best characterize spatial neighbors for accurate approximation of geometry occupancy probability and attribute intensity. Experiments demonstrate that DeepPCC remarkably outperforms state-of-the-art rules-based MPEG G-PCC and learning-based solutions both quantitatively and qualitatively, providing strong evidence that DeepPCC is a promising solution for emerging AI-based PCC.