← Back to all news
arXiv Machine Learning By Quanling Zhao, Ava Emami, Tsuyoshi Id\'e, Rajiv Joshi, Steven Holmes, Michael Passow, Tajana Rosing

Online Open-World Wafer Defect Discovery using Hyperdimensional Computing

Read the original on arXiv Machine Learning →

The paper introduces WaferHDC, an online open‑world wafer defect discovery framework that uses hyperdimensional computing to detect and group new defect signatures without retraining. It models normal spatial distributions and dynamically creates persistent defect memory, scaling with new defect modes. On the WM‑811K dataset, WaferHDC outperforms the strongest online baseline by 13.3 p.p. in adjusted Rand index and 10.5 p.p. in normalized mutual information.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

Related stories

arXiv Computer Vision
Sep 21

Morphology-Aware Ambiguity Learning for Wafer Defect Decision Support

The paper introduces a morphology‑aware ambiguity learning framework for wafer defect decision support. It enables three diagnostic actions—automatic single‑class diagnosis, assisted diagnosis with two plausible defect classes, and full review—by constructing a class‑level ambiguity matrix from wafer map characteristics. Experiments on the WM‑811K dataset demonstrate that the framework outperforms traditional methods, offering meaningful two‑class alternatives and reserving full review for truly ambiguous cases, with consistent performance across different backbone architectures.

By Seungjun Chu, Seokhyun Chung
More like this →
arXiv Machine Learning
Oct 8

Efficient Patch-Based Anomaly Detection Fused with Diffusion Driven Generative Modeling for Semiconductor Wafer Bin Map Open Set Anomaly Detection

The paper introduces a hybrid one‑class anomaly detector for semiconductor wafer bin maps that combines a patch‑based student‑teacher model (EfficientAD) with a denoising diffusion probabilistic model (DDPM). Trained on only 700 normal wafers, the fused system achieves an AUROC of 0.9985 on 18,658 held‑out wafers, reducing misclassifications from 852 (DDPM) and 1,412 (EfficientAD) to 618, with statistically significant improvements. Analysis reveals that the two modules make weakly overlapping errors, and that fixed‑weight fusion captures only 40‑70 % of the potential correction, highlighting the need for sample‑adaptive fusion and imbalance‑aware evaluation.

By Limon Bin Hossain, Md Sadib Rahman Ananta
diffusionbenchmarks
More like this →
arXiv AI
Sep 1

SePArate: Segmenting Patterns from Defects in Wafer Manufacturing Using Weak Supervision

SePArate is a weakly supervised method for segmenting patterns from defects in semiconductor wafers. It uses only image‑level annotations and follows a three‑phase training process: encoder pretraining, knowledge transfer for spatial cues, and training on synthetic mixed‑defect data. Experiments show that SePArate outperforms baseline approaches in pixel‑level defect segmentation.

By Dain Kwon, Changmin Shin, Sunjong Park, Kanghyun Choi, Hyeyoon Lee, Jaewon Jang, Minseok Choi, Jinho Lee
computer-vision
More like this →
arXiv Computer Vision
Sep 4

ISP-AD: A Large-Scale Real-World Dataset for Advancing Industrial Anomaly Detection with Synthetic and Real Defects

The paper introduces ISP-AD, the largest publicly available industrial anomaly detection dataset, featuring both synthetic and real defects from a factory floor. It focuses on challenging, small, weakly contrasted surface defects within highly variable structured patterns, addressing the bias of existing datasets toward ideal imaging conditions. Experiments demonstrate that even a small amount of weakly labeled real defects improves model generalization and that synthetic defects can serve as a useful cold‑start baseline for scalable training.

By Paul J. Krassnig, Dieter P. Gruber
benchmarks
More like this →