SFPNet: Sparse Focal Point Network for Semantic Segmentation on General LiDAR Point Clouds
This addresses the problem of model adaptability for robotic applications using diverse LiDAR types, representing an incremental improvement with a novel method for a known bottleneck.
The paper tackles the limited generalizability of LiDAR semantic segmentation methods across different LiDAR technologies by proposing SFPNet, a framework that replaces window-attention with sparse focal point modulation, achieving state-of-the-art results on solid-state LiDAR benchmarks and outperforming existing methods on a new hybrid-solid LiDAR dataset.
Although LiDAR semantic segmentation advances rapidly, state-of-the-art methods often incorporate specifically designed inductive bias derived from benchmarks originating from mechanical spinning LiDAR. This can limit model generalizability to other kinds of LiDAR technologies and make hyperparameter tuning more complex. To tackle these issues, we propose a generalized framework to accommodate various types of LiDAR prevalent in the market by replacing window-attention with our sparse focal point modulation. Our SFPNet is capable of extracting multi-level contexts and dynamically aggregating them using a gate mechanism. By implementing a channel-wise information query, features that incorporate both local and global contexts are encoded. We also introduce a novel large-scale hybrid-solid LiDAR semantic segmentation dataset for robotic applications. SFPNet demonstrates competitive performance on conventional benchmarks derived from mechanical spinning LiDAR, while achieving state-of-the-art results on benchmark derived from solid-state LiDAR. Additionally, it outperforms existing methods on our novel dataset sourced from hybrid-solid LiDAR. Code and dataset are available at https://github.com/Cavendish518/SFPNet and https://www.semanticindustry.top.