Transferring Styles for Reduced Texture Bias and Improved Robustness in Semantic Segmentation Networks
Not on the map. Why?
Abstract (arXiv)
Recent research has investigated the shape and texture biases of deep neural networks (DNNs) in image classification which influence their generalization capabilities and robustness. It has been shown that, in comparison to regular DNN training, training with stylized images reduces texture biases in image classification and improves robustness with respect to image corruptions. In an effort to advance this line of research, we examine whether style transfer can likewise deliver these two effects in semantic segmentation. To this end, we perform style transfer with style varying across artificial image areas. Those random areas are formed by a chosen number of Voronoi cells. The resulting style-transferred data is then used to train semantic segmentation DNNs with the objective of reducing their dependence on texture cues while enhancing their reliance on shape-based features. In our experiments, it turns out that in semantic segmentation, style transfer augmentation reduces texture bias and strongly increases robustness with respect to common image corruptions as well as adversarial attacks. These observations hold for convolutional neural networks and transformer architectures on the Cityscapes dataset as well as on PASCAL Context, showing the generality of the proposed method.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 37 of its 62 references that are in the map. First posted on arXiv 14 Jul 2025.
Builds on
37 of 62 references.
References · 37
- 2015Deep Residual Learning for Image Recognitioncited by 52,096
- 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
- 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
- 2023Segment Anythingcited by 8,234
- 2014Explaining and Harnessing Adversarial Examplescited by 7,364
- 2016Image-to-Image Translation with Conditional Adversarial Networkscited by 4,527
- 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,444
- 2018Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentationcited by 3,223
- 2021SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformerscited by 2,940
- 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,877
- 2016Perceptual Losses for Real-Time Style Transfer and Super-Resolutioncited by 2,694
- 2017Arbitrary Style Transfer in Real-time with Adaptive Instance Normalizationcited by 2,010
25 more
- 2018ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustnesscited by 1,432
- 2017StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translationcited by 1,039
- 2018Multimodal Unsupervised Image-to-Image Translationcited by 795
- 2021Intriguing Properties of Vision Transformerscited by 446
- 2019Approximating CNNs with Bag-of-local-Features models works surprisingly well on ImageNetcited by 260
- 2018Generalisation in humans and deep neural networkscited by 237
- 2021Partial success in closing the gap between human and machine visioncited by 185
- 2016Texture Networks: Feed-forward Synthesis of Textures and Stylized Imagescited by 174
- 2017On the Robustness of Semantic Segmentation Models to Adversarial Attackscited by 126
- 2019Benchmarking the Robustness of Semantic Segmentation Modelscited by 111
- 2021Are Convolutional Neural Networks or Transformers more like human vision?cited by 109
- 2017Exploring the structure of a real-time, arbitrary neural artistic stylization networkcited by 89
- 2021Delving Deep into the Generalization of Vision Transformers under Distribution Shiftscited by 85
- 2018Style Augmentation: Data Augmentation via Style Randomizationcited by 79
- 2020Shape-Texture Debiased Neural Network Trainingcited by 76
- 2019Interpreting Adversarially Trained Convolutional Neural Networkscited by 75
- 2020Dynamic Divide-and-Conquer Adversarial Training for Robust Semantic Segmentationcited by 38
- 2021Does enhanced shape bias improve neural network robustness to common corruptions?cited by 32
- 2024Reducing Texture Bias of Deep Neural Networks via Edge Enhancing Diffusioncited by 11
- 2023Understanding Segment Anything Model: SAM is Biased Towards Texture Rather than Shapecited by 10
- 2024Trapped in texture bias? A large scale comparison of deep instance segmentationcited by 10
- 2023Uncertainty-weighted Loss Functions for Improved Adversarial Attacks on Semantic Segmentationcited by 9
- 2023Uncertainty-based Detection of Adversarial Attacks in Semantic Segmentationcited by 5
- 2025Shape Bias and Robustness Evaluation via Cue Decomposition for Image Classification and Segmentationcited by 3
- 2021Content Preserving Image Translation with Texture Co-occurrence and Spatial Self-Similarity for Texture Debiasing and Domain Adaptationcited by 2
Built on by
Citing papers per year: 2026: 1.
1 paper.