Article Details

Abstract

Under low-light conditions in nighttime road scenarios, image features are susceptible to distortion, leading to a substantial decline in the accuracy of traditional semantic segmentation methods when processing such images—thus failing to meet the visual perception requirements of autonomous driving. To address this challenge, this study proposes a deep learning-based semantic segmentation approach for nighttime roads. Specifically, a teacher-student model architecture is first constructed, where the teacher model generates high-quality pseudo-labels to guide the online learning of the student model. Subsequently, a visibility enhancement module is employed to improve the quality of target images. Finally, a logit constraint learning module is introduced to mitigate the overconfidence issue of pseudo-labels induced by the traditional cross-entropy loss. Experimental results on the Cityscapes and Dark Zurich datasets demonstrate that the proposed method achieves favorable segmentation accuracy.

Keywords

semantic segmentation; autonomous driving; Deep Learning

References

1. [1] Xu Hui, Zhu Yuhua, Zhen Tong, etc A review of deep neural network image semantic segmentation methods [J]. Computer Science and Exploration, 2021, 15 (1): 47-59
2. [2] Li Canlin, Zhang Wenjiao, Shao Zhiwen, etc A semantic segmentation method for nighttime road scenes based on Trans nightSeg [J]. Journal of Zhejiang University (Engineering Edition), 2024, 58 (2): 294-303
3. [3] Shu Tao, Huang Jinsong Fine segmentation method for subordinate semantic channels in road scenes [J]. Journal of Navigation and Localization, 2022, 10 (5): 91-96
4. [4] Zhou Huaping, Li Meiguang Night time image semantic segmentation based on image style adversarial and dual category optimization [J]. Journal of Hubei University for Nationalities (Natural Science Edition), 2024, 42 (1): 69-74
5. [5] Krizhevsky A，Sutskever I，Hinton G E. Image Net Classification with DeepConvolutional Neural Networks [J].IEEE Transactions on Neural Information Processing Systems，2012，60（2）：1097-1105.
6. [6] Zhao H，Shi J，Qi X，et al. Pyramid Scene Parsing Network [C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，IEEE，2017：2881-2890.
7. [7] Chen L C，Zhu Y，Papandreou G，et al. Encoder-Decoder With Atrous Separable Convolution for Semantic Image Segmentation [C]//The European Conference on Computer Vision.Springer，2018：801-818.
8. [8] Chollet F. Xception：Deep Learning With Depthwise Separable Convolutions [C]//IEEE Conference on Computer Vision and Pattern Recognition，2017：1800-1807.
9. [9] Huang G，Liu Z，Van Der Maaten L，et al. Densely Connected Convolutional Networks [C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2017：4700-4708.
10. [10] Yang M，Yu K，Zhang C，et al. DenseASPP for Semantic Segmentation in Street Scenes [C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2018：3684-3692.
11. [11] Yu C，Wang J，Peng C，et al. BiSeNet: Bilateral Segmentation Network for Real-Time Semantic Segmentation [C]//Proceedings of the European Conference on Computer Vision （ECCV），2018：325-341.
12. [12] Zhao H，Zhang Y，Liu S，et al. PSANet：Point-Wise Spatial Attention Network for Scene Parsing [C]//Proceedings of the European Conference on Computer Vision （ECCV）.Springer，2018：267-283.
13. [13] Fu J，Liu J，Tian H，et al. Dual Attention Network for Scene Segmentation [C]//IEEE Conference on Computer Vision and Pattern Recognition.IEEE，2019：3141-3149.
14. [14] Wu Y，Pan C，Wang G，et al. Learning Semantic-Aware Knowledge Guidance for Low-Light Image Enhancement [C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.IEEE，2023：1662-1671.
15. [15] Zhao Zhengpeng, Li Jungang, Pu Yuanyuan Retinex Low Light Image Enhancement Based on Convolutional Neural Networks [J]. Computer Science, 2022, 49 (6): 199-209

Funding

Zhengzhou Sias College Campus level Research Project (2025XKE098)

About the authors

Wang Zhaoxiang (1999-), male, Han ethnicity, from Zhengzhou, Henan Province, China, teaching assistant, master's degree, research direction: machine vision.