cs.CVMar 15, 2026

Multi-Period Texture Contrast Enhancement for Low-Contrast Wafer Defect Detection and Segmentation

Authors: Zihan Zhang

Organizations: Engineering Technology Research Center of Henan Province for MEMS Manufacturing and Applications, School of Mechanics and Safety Engineering, Zhengzhou University, Zhengzhou 450001, China

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Wafer defect segmentation is pivotal for semiconductor yield optimization yet remains challenged by the intrinsic conflict between microscale anomalies and highly periodic, overwhelming background textures. Existing deep learning paradigms often falter due to feature dilution during downsampling and the lack of explicit mechanisms to disentangle low-contrast defects from process-induced noise. To transcend these limitations, we propose TexWDS, a texture-aware framework that harmonizes multi-scale feature retention with frequency-domain perturbation modeling. Our methodology incorporates three strategic innovations: (1) A Multi-scale Receptive Field Reweighting strategy is introduced to mitigate aliasing effects and preserve high-frequency details of micro-defects often lost in standard pyramidal architectures. (2) The Multi-scale Unified Semantic Enhancer (MUSE) integrates local appearance with global context encoding, effectively enhancing feature discriminability in low-visibility regions. (3) Crucially, we design a plug-and-play Multi-Periodic Texture Contrast Enhancement (MPTCE) module. By modeling texture disruptions in the frequency domain, MPTCE explicitly decouples non-periodic anomalies from structured backgrounds, boosting contrast for camouflaged defects. Extensive experiments on real-world industrial datasets demonstrate that TexWDS achieves a new state-of-the-art, surpassing the baseline by 8.3% in mAP50-95 and 7.7% in recall, while reducing the false positive rate by approximately 8.6%. These results underscore the framework's robustness in handling complex periodic patterns and its suitability for high-precision manufacturing inspection.

Figures & tables

Figure 1: Examples of scenarios where existing deep learning models perform suboptimally.
Figure 2: Normalized confusion matrix of YOLOv8-seg (baseline).
Figure 3: Schematic overview of the YOLOv8-seg (baseline) architecture.
Figure 4: Schematic overview of the proposed Texture-Aware Wafer Defect Segmentation (TexWDS) architecture, showing the integration positions of Stage 1 (P2 branch), Stage 2 (Multi-scale Unified Semantic Enhancer with Context Encoding (MUSE), and Stage 3 (Multi-Period Texture Contrast Enhancement, MPTCE) within the backbone–neck pipeline.
LayerInput SizeFilter SizeFilter NumStridePadding
Conv640×640640\times 6403×33\times 36421
Conv320×320320\times 3203×33\times 312821
C2f_MUSE160×160160\times 160\left[\begin{array}[]{c}1\times 1,64\\ 3\times 3,64\end{array}\right]\times 312811
Conv160×160160\times 1603×33\times 325621
C2f_MUSE80×8080\times 80\left[\begin{array}[]{c}1\times 1,128\\ 3\times 3,128\end{array}\right]\times 625611
Conv80×8080\times 803×33\times 351221
Table 1: The parameters of the proposed TexWDS model.
Figure 5: Comparison of feature maps from detection heads at different scales. (a), (b) and (c) represent the feature maps of the YOLOv8 model at the P2, P3, and P4 scales, respectively.
Figure 6: Illustration of the context structure within the proposed MUSE module, which contains three sub-branches: local, surrounding, and global channel attention.
Figure 7: Illustration of frequency decomposition, periodic disturbance extraction, and boundary-consistency attention within the proposed MPTCE module.
Figure 8: Normalized confusion matrix of the proposed Wafer defect detection framework, TexWDS.
P2 branchMUSEMPTCEGroupRecall ↑\uparrowPrecision ↑\uparrowmAP 50 ↑\uparrowmAP 50-95 ↑\uparrow
✗✗✗A071.271.472.747.9
✓\checkmark✗✗A174.773.874.950.9
✗✓\checkmark✗A274.676.275.750.3
✗✗✓\checkmarkA373.772.476.251.1
✓\checkmark✓\checkmark✗A477.176.776.952.9
✓\checkmark✗✓\checkmarkA575.877.576.751.4
Table 6: Ablation Study Results on the Contributions of the Three Main Components of the TexWDS Model: P2 Branch, MUSE, and MPTCE. (Note: ↑\uparrow indicates higher values are better.)
Figure 9: Comparison of feature activation maps in the final P3 detection layer between the baseline model and TexWDS. (a) Original image, (b) YOLOv8, and (c) TexWDS (Ours).
Figure 10: Visualization of periodic disturbance maps generated by the MPTCE module. (a) represents the original image, (b) represents the effect of overlaying the heatmap on the original image, and (c) represents the pure heatmap.
Figure 11: Comparison of detection results. (a), (b) and (c) represent the manually annotated defect bounding boxes and the detection results of YOLOv8 and TexWDS, respectively. Red boxes indicate detected defects, black boxes indicate detection failures.
Figure 12: Comparison of Segmentation Results. (a), (b) and (c) denote the manually annotated defect mask, the segmentation result of YOLOv8, and the segmentation result of TexWDS, respectively.
More in this topic

Image Segmentation

Image segmentation assigns pixels or voxels to objects and regions, supporting measurement, scene understanding, and decisions across image domains.

See all 399 papers

Explore similar work

CardsList
Aug 31, 2026

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

In semiconductor manufacturing, defect analysis is essential, but manual inspection cannot scale. However, existing automated inspection methods remain insufficient for root-cause analysis and process optimization. To this end, we present SePArate, a weakly supervised wafer defect segmentation method. SePArate enables pixel-level separation of patterns by leveraging only image-level annotations. It consists of a three-phase training: encoder pretraining, knowledge transfer to learn spatial cues, and training on synthetic mixed-defect data for accurate segmentation. Experiments demonstrate that SePArate outperforms the baselines.

Dain Kwon, Changmin Shin, Sunjong Park +5

Seoul National University · SK Hynix

Weakly Supervised Image SegmentationSemiconductor ManufacturingWeakly Supervised Semantic Segmentation
AskPDF
Oct 20, 2026

Online Open-World Wafer Defect Discovery using Hyperdimensional Computing

Conventional wafer defect analysis relies on offline, closed-world classifiers that fail to detect novel, unseen defect signatures emerging during continuous semiconductor manufacturing. To address this, we propose WaferHDC, an online open-world wafer defect discovery framework based on hyperdimensional computing (HDC). The memory-centric architecture of WaferHDC enables rapid anomaly detection and grouping of recurring failures into compact hypervector representations. This allows the memory space to dynamically scale with new defect modes without retraining a static classifier or archiving entire raw wafer history. WaferHDC first models normal spatial distributions, then dynamically groups recurring failures into persistent defect memory without labels or predefined class counts. It outperforms the strongest online baseline by 13.3 p.p. in adjusted Rand index and 10.5 p.p. in normalized mutual information on the industry-standard WM-811K dataset. These results demonstrate the effectiveness of WaferHDC for continuous and open-world wafer defect analysis.

Quanling Zhao, Ava Emami, Tsuyoshi Idé +4

University of California San Diego · IBM Research · Anderon

Online Anomaly DetectionIndustrial Anomaly DetectionAnomaly DetectionHyperdimensional Computing
AskPDF
Oct 7, 2026

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

Spatial defect signatures on wafer bin maps (WBMs) trace yield loss to specific process faults, yet supervised classifiers recognize only the defect types seen during training, and one-class detectors built on a single mechanism tend to capture either local structural deviations or global distributional violations, but rarely both. This work proposes a hybrid one-class framework that couples a patch-based student-teacher detector (EfficientAD) with a denoising diffusion probabilistic model (DDPM) used for partial-diffusion reconstruction, and fuses their percentile-calibrated scores through a fixed convex combination. Trained on only 700 normal wafers from the WM-38K mixed-type dataset and evaluated on 18,658 held-out wafers, the fused detector reached an AUROC of 0.9985 and reduced misclassifications from 852 (DDPM) and 1,412 (EfficientAD) to 618, with all pairwise differences significant at p < 0.001. Beyond aggregate accuracy, the analysis shows that the gain arises from weakly overlapping errors between the two modules, yet fixed-weight fusion recovers only 40-70% of the correction available to an oracle selector. Under the benchmark's inverted class balance, average precision and F1 saturate, while the Matthews correlation coefficient and negative predictive value expose unreliable normal predictions. Pixel-level maps further show that strong image-level separability does not imply spatial localization, and the diffusion module succeeds as a local density prior rather than through global geometric reasoning. These findings motivate sample-adaptive fusion and imbalance-aware evaluation of hybrid wafer anomaly detectors.

Limon Bin Hossain, Md Sadib Rahman Ananta

Department of Industrial and Production Engineering, Bangladesh University of Engineering and Technology (BUET), Dhaka, 1000, Bangladesh

Anomaly DetectionIndustrial Anomaly DetectionReconstruction-Based Anomaly Detection
AskPDF