Constrained Adaptive Projection with Pretrained Features for Anomaly Detection
Authors: Xingtai Gui, Di Wu, Yang Chang, Shicai Fan
IJCAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our method achieves state-of-the-art anomaly detection performance on semantic anomaly detection and sensory anomaly detection benchmarks including 96.5% AUROC on CIFAR100 dataset, 97.0% AUROC on CIFAR-10 dataset and 89.9% AUROC on Mv Tec dataset. extensive experiments and visualizations validate the effectiveness of our proposed framework for both sensory AD and semantic AD. |
| Researcher Affiliation | Academia | Xingtai Gui1 , Di Wu1 , Yang Chang1,2 , Shicai Fan1,2 1School of Automation Engineering, University of Electronic Science and Technology of China(UESTC) 2Shenzhen Institute for Advanced Study, UESTC |
| Pseudocode | No | The paper does not contain any clearly labeled pseudocode or algorithm blocks. The methods are described in textual paragraphs and a diagram. |
| Open Source Code | Yes | The source code of CAP is released at https://github.com/Tab Guigui/CAP. |
| Open Datasets | Yes | Four datasets including CIFAR-10, CIFAR-100, FMNIST and Mv Tec are used as the benchmarks. |
| Dataset Splits | No | The paper mentions training and testing phases and discusses results for different 'k' values, but it does not explicitly provide specific percentages or counts for train/validation/test splits in the main text. It refers to Appendix B for 'Detailed experiment configurations', but the main body lacks this information. |
| Hardware Specification | No | The paper mentions using 'Res Net152 and Wide Res Net50 pretrained on Image Net' but does not specify any hardware details such as GPU models, CPU types, or memory used for experiments. It references Appendix B for configurations but these details are not in the main text. |
| Software Dependencies | No | The paper does not list specific software dependencies with version numbers, such as programming languages, libraries (e.g., PyTorch, TensorFlow), or other tools used in the experimental setup. |
| Experiment Setup | No | The paper discusses parameters like 'k' and 'λ' in ablation studies, but it does not provide comprehensive experimental setup details such as learning rates, batch sizes, optimizers, or other training configurations in the main text. It states 'Detailed experiment configurations are shown in Appendix B.' |