针对沙漠地形特点，提出基于Transformer的十类语义分割模型，显著提升无人车导航精度。
DesertFormer: Transformer-Based Semantic Segmentation for Off-Road Desert Terrain Classification in Autonomous Navigation Systems
- 采用分层混合Transformer架构，适应沙漠低对比度与强光照变化场景
- 在4176张图像上训练，达到64.4% mIoU，较基线提升24.2个百分点
- 开源代码与可视化工具，适合自动驾驶与机器人路径规划研究者使用
可靠地形感知是无人车在非结构化野外环境自主导航的基础。沙漠地貌因各类地表间色度对比度低、光照极端变化、植被稀疏，难以满足传统道路分割模型假设。本文提出DesertFormer，基于SegFormer B2与分层混合Transformer（MiT-B2）骨干网络的语义分割系统，将地表划分为十类生态学意义明确的类别：树木、茂密灌木、干草、枯灌木、地面杂乱物、花朵、木头、岩石、地貌和天空，支持安全路径规划。在自建的4,176张512×512分辨率野外图像数据集上训练，mIoU达64.4%，像素准确率为86.1%，相较DeepLabV3 MobileNetV2基线（41.0% mIoU）提升24.2个百分点。进一步分析失败模式，发现地面杂乱物与地貌、干草与地貌易混淆，提出类别加权训练与复制粘贴增强策略以改善稀有类别表现。代码、模型权重及交互式推理仪表盘已公开于https://github.com/Yasaswini-ch/Vision-based-Desert-Terrain-Segmentation-using-SegFormer。
原文摘要 · Abstract (English)
Reliable terrain perception is a fundamental requirement for autonomous navigation in unstructured, off-road environments. Desert landscapes present unique challenges due to low chromatic contrast between terrain categories, extreme lighting variability, and sparse vegetation that defy the assumptions of standard road-scene segmentation models. We present DesertFormer, a semantic segmentation pipeline for off-road desert terrain analysis based on SegFormer B2 with a hierarchical Mix Transformer (MiT-B2) backbone. The system classifies terrain into ten ecologically meaningful categories -- Trees, Lush Bushes, Dry Grass, Dry Bushes, Ground Clutter, Flowers, Logs, Rocks, Landscape, and Sky -- enabling safety-aware path planning for ground robots and autonomous vehicles. Trained on a purpose-built dataset of 4,176 annotated off-road images at 512x512 resolution, DesertFormer achieves a mean Intersection-over-Union (mIoU) of 64.4% and pixel accuracy of 86.1%, representing a +24.2% absolute improvement over a DeepLabV3 MobileNetV2 baseline (41.0% mIoU). We further contribute a systematic failure analysis identifying the primary confusion patterns -- Ground Clutter to Landscape and Dry Grass to Landscape -- and propose class-weighted training and copy-paste augmentation for rare terrain categories. Code, checkpoints, and an interactive inference dashboard are released at https://github.com/Yasaswini-ch/Vision-based-Desert-Terrain-Segmentation-using-SegFormer.
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。