MDPI - Publisher of Open Access Journals

18 pages, 1514 KiB

Open AccessArticle

Contrastive Learning with Global and Local Representation for Mixed-Type Wafer Defect Recognition

by Shantong Yin, Yangkun Zhang and Rui Wang

Sensors 2025, 25(4), 1272; https://doi.org/10.3390/s25041272 - 19 Feb 2025

Viewed by 269

Recognizing defect patterns in semiconductor wafer bin maps (WBMs) poses a critical challenge in the integrated circuit (IC) manufacturing industry. The accurate classification and segmentation of these defect patterns are of utmost significance as they are key to tracing the root causes of defects, thereby reducing costs and enhancing both product efficiency and quality. As the manufacturing process grows in complexity, the WBM becomes intricate when multiple defect patterns coexist on a single wafer, making the recognition task increasingly complicated. In addition, traditional supervised learning methods require a large number of labeled samples, which is labor-intensive. In this paper, we present a self-supervised contrastive learning framework for the classification and segmentation of mixed-type WBM defect patterns. Our model incorporates a global module for contrastive learning that captures image-level representations, alongside a local module that targets the comprehension of regional details, which is helpful for the segmentation of defective patterns. Experimental results demonstrate that our model performs effectively in scenarios where there is a limited number of labeled examples and a wealth of unlabeled ones. Full article

(This article belongs to the Section Fault Diagnosis & Sensors)

► Show Figures

Figure 1

16 pages, 506 KiB

Open AccessArticle

Efficient Mixed-Type Wafer Defect Pattern Recognition Based on Light-Weight Neural Network

by Guangyuan Deng and Hongcheng Wang

Micromachines 2024, 15(7), 836; https://doi.org/10.3390/mi15070836 - 27 Jun 2024

Cited by 2 | Viewed by 1447

Abstract

Wafer defect pattern recognition can help engineers improve the production process of semiconductor chips. In real industrial scenarios, the recognition of mixed-type wafer defects is difficult and the production scale of semiconductor wafers is large, which requires high accuracy and speed in wafer defect pattern recognition. This study proposes a light-weight neural network model to efficiently recognize mixed-type wafer defects. The proposed model is constructed via inverted residual convolution blocks with attention mechanisms and large kernel convolution downsampling layers. The inference speed of the inverted residual convolution block is fast, and the attention mechanism can enhance feature extraction capabilities. Large kernel convolutions help the network retain more important feature information during downsampling operations. The experimental results on the real Mixed-type WM38 dataset show that the proposed model achieves a recognition accuracy of 98.69% with only 1.01 M parameters. Compared with some popular high-performance models and light-weight models, our model has advantages in both recognition accuracy and inference speed. Finally, we deploy the model as a TensorRT engine, which significantly improves the inference speed of the model, enabling it to process more than 1300 wafer maps per second. Full article

(This article belongs to the Special Issue Wide-Bandgap Semiconductor Devices: Materials, Fabrication, and Applications)

► Show Figures

Figure 1

17 pages, 9009 KiB

Open AccessArticle

Improved U-Net with Residual Attention Block for Mixed-Defect Wafer Maps

by Jaegyeong Cha and Jongpil Jeong

Appl. Sci. 2022, 12(4), 2209; https://doi.org/10.3390/app12042209 - 20 Feb 2022

Cited by 24 | Viewed by 4527

Abstract

Detecting defect patterns in semiconductors is very important for discovering the fundamental causes of production defects. In particular, because mixed defects have become more likely with the development of technology, finding them has become more complex than can be performed by conventional wafer defect detection. In this paper, we propose an improved U-Net model using a residual attention block that combines an attention mechanism with a residual block to segment a mixed defect. By using the proposed method, we can extract an improved feature map by suppressing irrelevant features and paying attention to the defect to be found. Experimental results show that the proposed model outperforms those in the existing studies. Full article

(This article belongs to the Special Issue Machine Learning in Manufacturing Technology and Systems)

► Show Figures

Figure 1

Search Results (3)

Further Information

Guidelines

MDPI Initiatives

Follow MDPI

Saved Queries

Search Filter Reset All

Years

Feature Papers

Subjects

Journals

Article Types

Countries / Regions

Search Results (3)

Further Information

Guidelines

MDPI Initiatives

Follow MDPI