# LiteViLNet: Lightweight Vision-LiDAR Fusion Network for Efficient Road Segmentation

arXiv: https://arxiv.org/abs/2605.21007 · HTML: https://mlsanity.com/paper/2605.21007/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 26 of its 37 references that are in the map.
- First posted on arXiv 20 May 2026.

## Builds on (most cited first, 26 of 26)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [Fully Convolutional Networks for Semantic Segmentation](https://mlsanity.com/paper/1411.4038/) (2014)
- [MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications](https://mlsanity.com/paper/1704.04861/) (2017)
- [Searching for MobileNetV3](https://mlsanity.com/paper/1905.02244/) (2019)
- [ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design](https://mlsanity.com/paper/1807.11164/) (2018)
- [BiSeNet: Bilateral Segmentation Network for Real-time Semantic Segmentation](https://mlsanity.com/paper/1808.00897/) (2018)
- [ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks](https://mlsanity.com/paper/1910.03151/) (2019)
- [InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions](https://mlsanity.com/paper/2211.05778/) (2022)
- [Cross-view Transformers for real-time Map-view Semantic Segmentation](https://mlsanity.com/paper/2205.02833/) (2022)
- [Coordinate Attention for Efficient Mobile Network Design](https://mlsanity.com/paper/2103.02907/) (2021)
- [CMX: Cross-Modal Fusion for RGB-X Semantic Segmentation with Transformers](https://mlsanity.com/paper/2203.04838/) (2022)
- [ESPNet: Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation](https://mlsanity.com/paper/1803.06815/) (2018)
- [PIDNet: A Real-time Semantic Segmentation Network Inspired by PID Controllers](https://mlsanity.com/paper/2206.02066/) (2022)
- [SNE-RoadSeg: Incorporating Surface Normal Information into Semantic Segmentation for Accurate Freespace Detection](https://mlsanity.com/paper/2008.11351/) (2020)
- [Progressive LiDAR Adaptation for Road Detection](https://mlsanity.com/paper/1904.01206/) (2019)
- [ORFD: A Dataset and Benchmark for Off-Road Freespace Detection](https://mlsanity.com/paper/2206.09907/) (2022)
- [RoadFormer: Duplex Transformer for RGB-Normal Semantic Road Scene Parsing](https://mlsanity.com/paper/2309.10356/) (2023)
- [Fast Road Segmentation via Uncertainty-aware Symmetric Network](https://mlsanity.com/paper/2203.04537/) (2022)
- [LOVON: Legged Open-Vocabulary Object Navigator](https://mlsanity.com/paper/2507.06747/) (2025)
- [SNE-RoadSegV2: Advancing Heterogeneous Feature Fusion and Fallibility Awareness for Freespace Detection](https://mlsanity.com/paper/2402.18918/) (2024)
- [CurbNet: Curb Detection Framework Based on LiDAR Point Cloud Segmentation](https://mlsanity.com/paper/2403.16794/) (2024)
- [TwinLiteNet+: An Enhanced Multi-Task Segmentation Model for Autonomous Driving](https://mlsanity.com/paper/2403.16958/) (2024)
- [ROD: RGB-Only Fast and Efficient Off-road Freespace Detection](https://mlsanity.com/paper/2508.08697/) (2025)
- [Annotation-Free Detection of Drivable Areas and Curbs Leveraging LiDAR Point Cloud Maps](https://mlsanity.com/paper/2603.27553/) (2026)
- [Annotation-Free Curb Detection Leveraging Altitude Difference Image](https://mlsanity.com/paper/2409.20171/) (2024)

## Built on by (most cited first, 1 of 1)

- [IAF-Net: Illumination-Adaptive Fusion for Low-Light Urban Road Segmentation](https://mlsanity.com/paper/2605.30939/) (2026)

## Abstract (arXiv)

Road segmentation is a fundamental perception task for autonomous driving and mobile robotics, where both appearance and geometric cues must be processed under edge-computing constraints. Existing multi-modal approaches often improve accuracy with large encoders or expensive global interaction, which limits their use on embedded platforms. We present LiteViLNet, a lightweight RGB-geometry fusion network that combines a MobileNetV3 RGB encoder with a 0.12M-parameter depth-wise-separable geometry encoder. A multi-scale feature fusion module performs modality-specific enhancement, global-query cross-modal interaction, and adaptive gating, while a depth-wise large-kernel bridge enlarges the contextual support of the deepest representation with low overhead. The resulting U-Net-style decoder uses deep supervision only during training. On the KITTI Road benchmark, the 14.04M-parameter full model obtains 97.23±0.15% MaxF. On the held-out ORFD test set under the released OFF-Net evaluation protocol, the full model achieves 96.74±0.09% F-score and 93.68±0.18% IoU. On a Jetson Orin NX, model-only PyTorch FP16 inference reaches 22.18±0.21 FPS; a separate TensorRT FP16 measurement reaches 68.73±0.06 FPS on the Jetson. Camera-depth adaptations and perception-and-control demonstrations on three heterogeneous robot platforms further illustrate the portability of the dual-stream design.