mlsanityData to 14 Oct 2026

Anomaly Discovery in Semantic Segmentation via Distillation Comparison Networks

arXiv 2112.09908 ↗First posted 18 Dec 2021cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

This paper aims to address the problem of anomaly discovery in semantic segmentation. Our key observation is that semantic classification plays a critical role in existing approaches, while the incorrectly classified pixels are easily regarded as anomalies. Such a phenomenon frequently appears and is rarely discussed, which significantly reduces the performance of anomaly discovery. To this end, we propose a novel Distillation Comparison Network (DiCNet). It comprises of a teacher branch which is a semantic segmentation network that removed the semantic classification head, and a student branch that is distilled from the teacher branch through a distribution distillation. We show that the distillation guarantees the semantic features of the two branches hold consistency in the known classes, while reflect inconsistency in the unknown class. Therefore, we leverage the semantic feature discrepancy between the two branches to discover the anomalies. DiCNet abandons the semantic classification head in the inference process, and hence significantly alleviates the issue caused by incorrect semantic classification. Extensive experimental results on StreetHazards dataset and BDD-Anomaly dataset are conducted to verify the superior performance of DiCNet. In particular, DiCNet obtains a 6.3% improvement in AUPR and a 5.2% improvement in FPR95 on StreetHazards dataset, achieves a 4.2% improvement in AUPR and a 6.8% improvement in FPR95 on BDD-Anomaly dataset. Codes are available at https://github.com/zhouhuan-hust/DiCNet.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 35 of its 40 references that are in the map. First posted on arXiv 18 Dec 2021.

Builds on

35 of 40 references.

References · 35

  1. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  2. 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,448
  3. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,926
  4. 2016DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFscited by 3,241
  5. 2016Pyramid Scene Parsing Networkcited by 2,479
  6. 2016A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networkscited by 2,385
  7. 2014FitNets: Hints for Thin Deep Netscited by 1,653
  8. 2015SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentationcited by 1,487
  9. 2019Self-training with Noisy Student improves ImageNet classificationcited by 1,311
  10. 2018BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learningcited by 1,288
  11. 2018Averaging Weights Leads to Wider Optima and Better Generalizationcited by 1,209
  12. 2017High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANscited by 1,179
23 more
  1. 2016Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfercited by 1,077
  2. 2017Selective Classification for Deep Neural Networkscited by 852
  3. 2017Deep Mutual Learningcited by 628
  4. 2017Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discoverycited by 566
  5. 2019A Simple Baseline for Bayesian Uncertainty in Deep Learningcited by 560
  6. 2019Similarity-Preserving Knowledge Distillationcited by 516
  7. 2018GANomaly: Semi-Supervised Anomaly Detection via Adversarial Trainingcited by 409
  8. 2019On the Efficacy of Knowledge Distillationcited by 381
  9. 2018To Trust Or Not To Trust A Classifiercited by 207
  10. 2019Addressing Failure Prediction by Learning Model Confidencecited by 186
  11. 2017Computer Vision for Autonomous Vehicles: Problems, Datasets and State of the Artcited by 182
  12. 2018Deep Autoencoding Models for Unsupervised Anomaly Segmentation in Brain MR Imagescited by 146
  13. 2018OCNet: Object Context Network for Scene Parsingcited by 140
  14. 2019Detecting the Unexpected via Image Resynthesiscited by 137
  15. 2018Out-of-Distribution Detection Using an Ensemble of Self Supervised Leave-out Classifierscited by 126
  16. 2021Pixel-wise Anomaly Detection in Complex Driving Scenescited by 109
  17. 2020Synthesize then Compare: Detecting Failures and Anomalies for Semantic Segmentationcited by 96
  18. 2018Knowledge Distillation in Generations: More Tolerant Teachers Educate Better Studentscited by 50
  19. 2019TRADI: Tracking deep neural network weight distributions for uncertainty estimationcited by 39
  20. 2017Limiting the Reconstruction Capability of Generative Neural Network using Negative Learningcited by 16
  21. 2021Detecting Anomalies in Semantic Segmentation with Prototypescited by 13
  22. 2019An Alarm System For Segmentation Algorithm Based On Shape Modelcited by 8
  23. 2020Gated Path Selection Network for Semantic Segmentationcited by 6

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2021Double Similarity Distillation for Semantic Image Segmentation
  2. 2026The Surprising Effectiveness of Canonical Knowledge Distillation for Semantic Segmentation
  3. 2022Distilling Inter-Class Distance for Semantic Segmentation
  4. 2023AICSD: Adaptive Inter-Class Similarity Distillation for Semantic Segmentation
  5. 2024VL4AD: Vision-Language Models Improve Pixel-wise Anomaly Detection
  6. 2024Relative Difficulty Distillation for Semantic Segmentation
  7. 2024I2CKD : Intra- and Inter-Class Knowledge Distillation for Semantic Segmentation
  8. 2024Dual-Student Knowledge Distillation Networks for Unsupervised Anomaly Detection