DiffSSC: Semantic LiDAR Scan Completion using Denoising Diffusion Probabilistic Models

2024-09-26

arXiv.org

AI Summary

To address the challenges of occluded region perception in autonomous driving caused by sparse and semantically impoverished LiDAR point clouds, this paper proposes the first decoupled conditional diffusion model for Semantic Scene Completion (SSC). Methodologically, it introduces denoising diffusion probabilistic models to SSC for the first time, innovatively decoupling the diffusion processes for geometric point-space reconstruction and semantic label prediction. A local-global regularization loss is designed to enhance generation stability, while multi-scale feature modeling and semantic consistency constraints are integrated. Evaluated on benchmark datasets—including SemanticKITTI and nuScenes—our approach achieves state-of-the-art performance across all three core metrics: completion completeness, geometric accuracy, and semantic consistency, significantly outperforming existing methods.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

Perception systems play a crucial role in autonomous driving, incorporating multiple sensors and corresponding computer vision algorithms. 3D LiDAR sensors are widely used to capture sparse point clouds of the vehicle's surroundings. However, such systems struggle to perceive occluded areas and gaps in the scene due to the sparsity of these point clouds and their lack of semantics. To address these challenges, Semantic Scene Completion (SSC) jointly predicts unobserved geometry and semantics in the scene given raw LiDAR measurements, aiming for a more complete scene representation. Building on promising results of diffusion models in image generation and super-resolution tasks, we propose their extension to SSC by implementing the noising and denoising diffusion processes in the point and semantic spaces individually. To control the generation, we employ semantic LiDAR point clouds as conditional input and design local and global regularization losses to stabilize the denoising process. We evaluate our approach on autonomous driving datasets and our approach outperforms the state-of-the-art for SSC.

Problem

• Completing sparse LiDAR scans with semantics for autonomous driving
• Predicting occluded geometry and semantics in 3D scenes
• Enhancing scene representation using diffusion models

Innovation

• Uses denoising diffusion for semantic completion
• Employs semantic LiDAR as conditional input
• Applies local and global regularization losses