SNE-RoadSegV2: Advancing Heterogeneous Feature Fusion and Fallibility Awareness for Freespace Detection
- 26 February 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-9
Abstract
Feature-fusion networks with duplex encoders have proven to be an effective technique to solve the road freespace detection problem. However, despite the compelling results achieved by previous research efforts, the exploration of adequate and discriminative heterogeneous feature fusion, as well as the development of fallibility-aware loss functions, remains relatively scarce. This article makes several significant contributions to address these limitations: 1) it presents a novel heterogeneous feature fusion block, comprising a holistic attention module, a heterogeneous feature contrast descriptor, and an affinity-weighted feature recalibrator, enabling a more in-depth exploitation of the inherent characteristics of the extracted features, 2) it incorporates both inter-scale and intra-scale skip connections into the decoder architecture, while eliminating redundant ones, leading to both improved accuracy and computational efficiency, and 3) it introduces two fallibility-aware loss functions that separately focus on semantic-transition and depth-inconsistent regions, collectively contributing to greater supervision during model training. Our proposed SNE-RoadSegV2, which incorporates all these innovative components, demonstrates superior performance in comparison to all other freespace detection algorithms across multiple public datasets.Keywords
All Related Versions
Funding Information
- National Natural Science Foundation of China (62473288)
- National Natural Science Foundation of China (62233013)
- Science and Technology Commission of Shanghai Municipal (22511104500)
- Fundamental Research Funds for the Central Universities, NIO University Program
- Xiaomi Young Talents Program
- Research Leading to these Results has also Received Partial Funding from European Commission-European Union (under HORIZON EUROPE (HORIZON Research and Innovation Actions) (101093003)
This publication has 50 references indexed in Scilit:
- Dual geometric perception for cross-domain road segmentationDisplays, 2022
- FANet: Feature aggregation network for RGBD saliency detectionSignal Processing: Image Communication, 2022
- Bi-directional Cross-Modality Feature Propagation with Separation-and-Aggregation Gate for RGB-D Semantic Segmentation2020
- Hierarchical Dynamic Filtering Network for RGB-D Salient Object Detection2020
- SNE-RoadSeg: Incorporating Surface Normal Information into Semantic Segmentation for Accurate Freespace Detection2020
- LIDAR–camera fusion for road detection using fully convolutional neural networksRobotics and Autonomous Systems, 2019
- Augmented Reality Meets Computer Vision: Efficient Data Generation for Urban Driving ScenesInternational Journal of Computer Vision, 2018
- FuseNet: Incorporating Depth into Semantic Segmentation via Fusion-Based CNN Architecture2017
- U-Net: Convolutional Networks for Biomedical Image Segmentation2015
- Road Scene Segmentation from a Single Image2012