Few-Shot Defect Image Generation via Defect-Aware Feature Manipulation
Authors: Yuxuan Duan, Yan Hong, Li Niu, Liqing Zhang
AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on MVTec AD dataset not only validate the effectiveness of our method in generating realistic and diverse defect images, but also manifest the benefits it brings to downstream defect inspection tasks. |
| Researcher Affiliation | Academia | Yuxuan Duan, Yan Hong, Li Niu , Liqing Zhang* Mo E Key Lab of Artificial Intelligence, Shanghai Jiao Tong University sjtudyx2016@sjtu.edu.cn, yanhong.sjtu@gmail.com, ustcnewly@sjtu.edu.cn, zhang-lq@cs.sjtu.edu.cn |
| Pseudocode | No | The paper describes the method using text and architectural diagrams (Figure 1, Figure 2) but does not include structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | Codes are available at https://github.com/Ldhlwh/DFMGAN. |
| Open Datasets | Yes | MVTec Anomaly Detection2 (MVTec AD) (Bergmann et al. 2019) is an open dataset containing ten object categories and five texture categories commonly seen, with up to eight defect categories for each object/texture category. All the images are accompanied with pixel-level masks showing the defect regions. Although originally designed for defect localization, MVTec AD fits the experimental setting for few-shot defect image generation since most object/texture categories have 200 400 defect-free samples, and most defect categories have 10 25 defect images. 2https://www.mvtec.com/company/research/datasets/mvtecad, released under CC BY-NC-SA 4.0. |
| Dataset Splits | No | The paper describes how the dataset is used for different stages and tasks (e.g., 'randomly choose 1/3 of the dataset images from each defect category as the base sets, and the other 2/3 from each category are combined as the test set' for defect classification), but does not explicitly provide percentages or counts for a train/validation/test split for the primary GAN training dataset. |
| Hardware Specification | No | The paper does not provide specific details regarding the hardware (e.g., GPU models, CPU types, memory) used for running the experiments. |
| Software Dependencies | No | The paper does not provide specific version numbers for software dependencies or libraries used in the experiments. |
| Experiment Setup | No | The paper states 'See supplementary material for implementation details' and does not provide specific hyperparameter values like learning rate, batch size, or number of epochs in the main text. |