SFPNet: Sparse Focal Point Network for Semantic Segmentation on General LiDAR Point Clouds

TL;DR

SFPNet introduces a flexible LiDAR semantic segmentation framework that generalizes across different LiDAR types, leveraging sparse focal point modulation to improve context extraction and achieve state-of-the-art results on multiple datasets.

Contribution

The paper proposes a novel sparse focal point modulation method and a hybrid-solid LiDAR dataset, enhancing generalizability and performance in LiDAR semantic segmentation.

Findings

Competitive performance on mechanical spinning LiDAR benchmarks.

State-of-the-art results on solid-state LiDAR benchmarks.

Outperforms existing methods on the new hybrid-solid LiDAR dataset.

Abstract

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.