DEQ-Det: Dual Evidence Queries for Few-Shot Analog IC Device Detection
Authors: Zidong Zhao, Rong Huang, Chengxing Fan, Jie Hu, Malik Haris, Ghazal A. Fahmy, Morteza Mousazadeh, Guoqiang Zhang
Organizations: Zhejiang University, China · Donghua University, China · National Telecommunication Institute, Egypt · Urmia University, Iran · Ningbo Tech University, China
Detecting analog integrated circuit (IC) devices in layout images supports circuit analysis and inspection. With only a few annotated examples, variation in device geometry and visual similarity to dummy or routing structures make reliable detection difficult. Foreground similarity alone can highlight plausible targets without distinguishing them from competing layout patterns. We propose Dual Evidence Queries for Detection (DEQ-Det), which constructs complementary foreground and background references from the same support images without additional annotations. Foreground references capture target morphology, while selected background references characterize surrounding layout patterns. Both sources guide query selection and region verification, combining target appearance with layout context. Across five held-out device categories and three random seeds, DEQ-Det achieves macro AP50 of 39.41 and 53.03 at 10 and 30 shots, exceeding the strongest listed baselines by 3.97 and 9.71 points, respectively. A matched 30-shot ablation further shows that RoI background evidence reduces pooled false positives at comparable recall, supporting its role in interference suppression.
Figures & tables
Fig. 1: Analog layout variability and ambiguity. Device geometries vary within categories; non-target dummy structures share outlines or repeated patterns with devices. Blue/green boxes indicate devices, and the red dashed region indicates dummies.
Fig. 2: DEQ-Det. Foreground references guide target selection, and filtered background references supply layout context for interference suppression. Their means guide query selection, while individual bank records enter the RoI relation head. Foreground probability and predicted localization quality jointly rank candidate detections.
Fig. 3: Qualitative comparison at 30 shots on Diode and NPN/PNP examples. Columns show ground truth, TFA, DE-ViT, FSOD-VFM, and DEQ-Det. Cyan dashed boxes denote ground truth; green/red boxes denote matched/unmatched predictions at IoU 0.5.
Method
10-shot
30-shot
Template matching
13.20 ± 2.00
17.17 ± 2.75
DINOv2 prototype
12.92 ± 1.43
20.80 ± 3.28
TFA [ 5 ]
18.71 ± 1.56
21.93 ± 1.08
Meta R-CNN [ 6 ]
31.02 ± 4.81
43.32 ± 5.78
FSCE [ 7 ]
32.38 ± 2.17
38.65 ± 0.35
VFA [ 8 ]
32.89 ± 1.98
38.01 ± 2.18
TABLE I: Five-category macro AP50 at 10 and 30 shots, mean ± sample standard deviation over seeds 1/2/3.
Fig. 4: Matched 30-shot component ablations: five-category macro AP50, mean and sample standard deviation over three seeds. Blue/orange bars denote full/ablated models.
Decision models can answer a typed question about an image with a probability per option, without generating text. In our case study of Clef-flash (post-trained Qwen3.5-9B backbone) on industrial inspection benchmarks, we ask whether it localizes the anomalies it detects. We obtained uneven results with zero-shot detection: per-category image AUROC 0.924 on MVTec AD and 0.668 on MVTec AD 2. On the former, the decision head's routing attention produces 0.391 mean per-image pixel AUROC, and hiding the image tokens from it leaves detection unchanged. The multimodal LLM's backbone attention does point at the defect, however it falls below a center prior when we change question. It falls below chance if we remove textual context in the question, partly due to backbone attention leaving the object, while the head's map hardly changes. The backbone's states carry the location when probed with normal images, on the test-split images, however, the decision depends on the defect region. A larger model, Clef, that comes with a post-trained Qwen3.8-27B backbone, gives no useful map in the decision head either. We found that it's possible to recover anomaly location by asking question specifically on an image region, retraining the head or reading the backbone, but none matches dedicated few-shot localizers.
The integration of artificial intelligence into computer-aided design frameworks has sparked a shift in the design of analog integrated circuits (ICs), transitioning the field from using manual and algorithmic-based solutions to adopting automated and intelligent paradigms. In this scenario, the GDSII file represents the industry-standard database containing the ultimate and most accurate source of information of the analog circuit, encapsulating the complex physical geometries and parasitic realities that define tape out performance. This paper proposes THEIA, a novel dataset containing thousands of layout images paired with question-answer conversations, along with a benchmark that employs a fine-tuned vision-language model (VLM) to analyze GDSII files of analog circuits, enabling designers to interact with and query physical layouts as intuitive, meaningful entities. Experimental results using thousands of analog designs across five realistic tasks demonstrate that the proposed fine-tuned VLM outperforms state-of-the-art general-purpose VLMs by a significant margin (up to 73%), highlighting a fundamental gap between general-purpose multimodal reasoning and domain-specific layout understanding.
Giuseppe Chiari, Michele Piccoli, Federico Viola +1
Multimodal large language models have demonstrated strong defect recognition capability in industrial anomaly detection. However, in lithography review, merely determining whether an image contains a defect is insufficient for engineering inspection; models must also understand defect morphology, spatial location, and the potential causes supported by visible evidence. To this end, this paper proposes LDU-Bench, a multi-task multimodal benchmark for lithography defect understanding. Constructed from real lithography and integrated-circuit review images, LDU-Bench decomposes the review workflow into four independent tasks: defect triage, morphology recognition, coarse localization, and image-conditioned cause analysis. It systematically evaluates models using task-level metrics, diagnostic readouts, and the Lithography Closure Score (LCS). Experimental results show that although existing MLLMs can perform defect triage relatively reliably, this ability does not stably transfer to downstream review stages. Morphology alignment, effective localization, and evidence-to-cause mapping remain the major bottlenecks. Further diagnostics indicate that this capability break is not a fluctuation of a single metric, but reflects insufficient structured understanding across semantic levels. Overall, LDU-Bench provides a quantifiable and diagnostic unified platform for evaluating the usability, failure points, and capability boundaries of industrial MLLMs in lithography review chains.