Defect-GAN: High-Fidelity Defect Synthesis for Automated Defect Inspection

Zhang, Gongjie; Cui, Kaiwen; Hung, Tzu-Yi; Lu, Shijian

doi:10.1109/wacv48630.2021.00257

Cited by 88 publications

(46 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Why is GenCo effective? In data-limited image generation, one major issue is that discriminator in GANs tends to suffer from over-fitting by capturing simple structures and patterns only (Bau et al 2019;Zhang et al 2021). The proposed GenCo mitigates this issue by co-training two discriminators in WeCo and DaCo.…”

Section: Overall Training Objectivementioning

confidence: 99%

GenCo: Generative Co-training for Generative Adversarial Networks with Limited Data

Cui

Huang

Luo

et al. 2022

AAAI

Self Cite

View full text Add to dashboard Cite

Training effective Generative Adversarial Networks (GANs) requires large amounts of training data, without which the trained models are usually sub-optimal with discriminator over-fitting. Several prior studies address this issue by expanding the distribution of the limited training data via massive and hand-crafted data augmentation. We handle data-limited image generation from a very different perspective. Specifically, we design GenCo, a Generative Co-training network that mitigates the discriminator over-fitting issue by introducing multiple complementary discriminators that provide diverse supervision from multiple distinctive views in training. We instantiate the idea of GenCo in two ways. The first way is Weight-Discrepancy Co-training (WeCo) which co-trains multiple distinctive discriminators by diversifying their parameters. The second way is Data-Discrepancy Co-training (DaCo) which achieves co-training by feeding discriminators with different views of the input images. Extensive experiments over multiple benchmarks show that GenCo achieves superior generation with limited training data. In addition, GenCo also complements the augmentation approach with consistent and clear performance gains when combined.

show abstract

Section: Overall Training Objectivementioning

confidence: 99%

GenCo: Generative Co-training for Generative Adversarial Networks with Limited Data

Cui

Huang

Luo

et al. 2022

AAAI

Self Cite

View full text Add to dashboard Cite

show abstract

“…The synthetic anomalies are subsequently used in conjunction with readily available, defect-free images to either fine-tune supervised methods [20] or train them from scratch [18]. While defect/anomaly synthesis by means of GANs is also used to improve performance at general surface inspection tasks [31,32], GANs are known to be notoriously difficult to train [33], diminishing the applicability of the developed solutions.…”

Section: Post Hoc Adaptation Techniquesmentioning

confidence: 99%

Increasing the Generalization of Supervised Fabric Anomaly Detection Methods to Unseen Fabrics

Rippel

Zwinge

Merhof

2022

Sensors

View full text Add to dashboard Cite

Fabric ad tries to detect anomalies (i.e., defects) in fabrics, and fabric ad approaches are continuously improved with respect to their ad performance. However, developed solutions are known to generalize poorly to previously unseen fabrics, posing a crucial limitation to their applicability. Moreover, current research focuses on adapting converged models to previously unseen fabrics in a post hoc manner, rather than training models that generalize better in the first place. In our work, we explore this potential for the first time. Specifically, we propose that previously unseen fabrics can be regarded as shifts in the underlying data distribution. We therefore argue that factors which reportedly improve a model’s resistance to distribution shifts should also improve the performance of supervised fabric ad methods on unseen fabrics. Hence, we assess the potential benefits of: (I) vrm techniques adapted to the fabric ad use-case, (II) different loss functions, (III) ImageNet pre-training, (IV) dataset diversity, and (V) model architecture as well as model complexity. The subsequently performed large-scale analysis reveals that (I) only the vrm technique, AugMix, consistently improves performance on unseen fabrics; (II) hsc outperforms other loss functions when combined with AugMix and (III) ImageNet pre-training, which is already beneficial on its own; (IV) increasing dataset diversity improves performance on unseen fabrics; and (V) architectures with better ImageNet performance also perform better on unseen fabrics, yet the same does not hold for more complex models. Notably, the results show that not all factors and techniques which reportedly improve a model’s resistance to distribution shifts in natural images also improve the generalization of supervised fabric ad methods to unseen fabrics, demonstrating the necessity of our work. Additionally, we also assess whether the performance gains of models which generalize better propagate to post hoc adaptation methods and show this to be the case. Since no suitable fabric dataset was publicly available at the time of this work, we acquired our own fabric dataset, called OLP, as the basis for the above experiments. OLP consists of 38 complex, patterned fabrics, more than 6400 images in total, and is made publicly available.

show abstract

“…Ö Commutator [25] Ö Chiller [26] Ö Ö Fiber layer up [27] Ö Laser welding [28] Ö General methods [29,37] Ö Ö…”

Section: -4 / S Mou • Invited Paper Sid 2022 Digest • 975unclassified

72‐4: Invited Paper: Synthetic Defect Generation for Display Front‐of‐Screen Quality Inspection: A Survey

Mou

Cao

Hong

et al. 2022

Symp Digest of Tech Papers

View full text Add to dashboard Cite

Display front‐of‐screen (FOS) quality inspection is essential for the mass production of displays in the manufacturing process. However, the severe imbalanced data, especially the limited number of defective samples, has been a long‐standing problem that hinders the successful application of deep learning algorithms. Synthetic defect data generation can help address this issue. This paper reviews the state‐of‐the‐art synthetic data generation methods and the evaluation metrics that can potentially be applied to display FOS quality inspection tasks.

show abstract

Defect-GAN: High-Fidelity Defect Synthesis for Automated Defect Inspection

Cited by 88 publications

References 56 publications

GenCo: Generative Co-training for Generative Adversarial Networks with Limited Data

GenCo: Generative Co-training for Generative Adversarial Networks with Limited Data

Increasing the Generalization of Supervised Fabric Anomaly Detection Methods to Unseen Fabrics

72‐4: Invited Paper: Synthetic Defect Generation for Display Front‐of‐Screen Quality Inspection: A Survey

Contact Info

Product

Resources

About