轻量级视觉主导场景补全，用蒸馏技术提升精度同时降耗。
L2COcc: Lightweight Camera-Centric Semantic Scene Completion via Distillation of LiDAR Model
- 通过高效体素变换器与跨模态知识蒸馏，实现低开销视觉场景补全。
- 在SemanticKITTI和SSCBench-KITTI-360上均超越当前最优视觉方法。
- 模型内存与推理时间减少超23%，适合边缘部署的自动驾驶系统。
语义场景补全（SSC）是自动驾驶感知系统的关键环节，旨在从传感器数据中推断场景的3D语义占据情况。以往研究依赖大量计算密集型和内存占用高的3D操作，显著增加平台训练与推理负担。本文提出L2COcc，一种轻量级、以摄像头为中心的SSC框架，同时支持激光雷达输入。通过提出的高效体素变压器（EVT）及跨模态知识模块——特征相似性蒸馏（FSD）、TPV蒸馏（TPVD）和预测对齐蒸馏（PAD），本方法大幅降低计算开销，同时保持高精度。实验表明，该方法在SemanticKITTI与SSCBench-KITTI-360基准上均优于现有纯视觉方法。此外，相比当前最优方法，模型内存消耗与推理时间均减少超过23%。代码已公开于：https://studyingfufu.github.io/L2COcc/。
原文摘要 · Abstract (English)
Semantic Scene Completion (SSC) constitutes a pivotal element in autonomous driving perception systems, tasked with inferring the 3D semantic occupancy of a scene from sensory data. To improve accuracy, prior research has implemented various computationally demanding and memory-intensive 3D operations, imposing significant computational requirements on the platform during training and testing. This paper proposes L2COcc, a lightweight camera-centric SSC framework that also accommodates LiDAR inputs. With our proposed efficient voxel transformer (EVT) and cross-modal knowledge modules, including feature similarity distillation (FSD), TPV distillation (TPVD) and prediction alignment distillation (PAD), our method substantially reduce computational burden while maintaining high accuracy. The experimental evaluations demonstrate that our proposed method surpasses the current state-of-the-art vision-based SSC methods regarding accuracy on both the SemanticKITTI and SSCBench-KITTI-360 benchmarks, respectively. Additionally, our method is more lightweight, exhibiting a reduction in both memory consumption and inference time by over 23% compared to the current state-of-the-arts method. Code is available at our project page:https://studyingfufu.github.io/L2COcc/.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。