Nan Wang

Prof. Nan Wang


Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
23 December 2020 in PLOS ONE
Verified
Journal Article
Open access
PDF file
WJWei Jiang
Wei Jiang
KZKai Zhang
Kai Zhang
Nan WangNan Wang
Nan Wang
MYMiao Yu
Miao Yu
To solve overfitting in machine learning, we propose a novel data augmentation method called MeshCut, which uses a mesh-like mask to segment the whole image to achieve more partial diversified information. In our experiments, this strategy outperformed the existing augmentation strategies and achieved state-of-the-art results in a variety of computer vision tasks. MeshCut is also an easy-to-implement strategy that can efficiently improve the performance of the existing convolutional neural network models by a good margin without careful hand-tuning. The performance of such a strategy can be further improved by incorporating it into other augmentation strategies, which can make MeshCut a promising baseline strategy for future data augmentation algorithms.

25 February 2021 in Optics Letters
Verified
Journal Article
Closed access
PDF file
Nan WangNan Wang
Nan Wang
WJWei Jiang
Wei Jiang
YZYu Zhang
Yu Zhang
In lithography, misalignment measurement with a large range and high precision in two dimensions for the overlay is a fundamental but challenging problem. For moiré-based misalignment measurement schemes, one potential solution is considered to be the use of circular gratings, whose formed moiré fringes are symmetric, isotropic, and aperiodic. However, due to the absence of proper analytical arithmetic, the measurement accuracy of such schemes is in the tens of nanometers, resulting in their application being limited to only coarse alignments. To cope with this problem, we propose a novel deep learning–based misalignment measurement strategy inspired by deep convolutional neural networks. The experimental results show that the proposed scheme can achieve nanoscale accuracy with micron-scale circular alignment marks. Relative to the existing strategies, this strategy has much higher precision in misalignment measurement and much better robustness to fabrication defects and random noise. This enables a one-step two-dimensional nanoscale alignment scheme for proximity, x-ray, extreme ultraviolet, projective, and nanoimprint lithographies.

16 August 2022 in Journal of Sensors
Verified
Journal Article
Open access
PDF file
ZWZhixue Wang
Zhixue Wang
ZYZhang Yu
Zhang Yu
LLLin Luo
Lin Luo
Nan WangNan Wang
Nan Wang
This paper proposed a novel anomaly detection (AD) approach of high-speed train images based on convolutional neural networks and the Vision Transformer. Different from previous AD works, in which anomalies are identified with a single image using classification, segmentation, or object detection methods, the proposed method detects abnormal difference between two images taken at different times of the same region. In other words, we cast anomaly detection problem with a single image into a difference detection problem with two images. The core idea of the proposed method is that the “anomaly” commonly represents an abnormal state instead of a specific object, and this state should be identified by a pair of images. In addition, we introduced a deep feature difference AD network (AnoDFDNet) which sufficiently explored the potential of the Vision Transformer and convolutional neural networks. To verify the effectiveness of the proposed AnoDFDNet, we gathered three datasets, a difference dataset (Diff dataset), a foreign body dataset (FB dataset), and an oil leakage dataset (OL dataset). Experimental results on the above datasets demonstrate the superiority of the proposed method. In terms of the F1-score, the AnoDFDNet obtained 76.24%, 81.04%, and 83.92% on Diff dataset, FB dataset, and OL dataset, respectively.

1 August 2021 in Optics and Lasers in Engineering
Verified
Journal Article
Closed access
Nan WangNan Wang
Nan Wang
WJWei Jiang
Wei Jiang
YZYu Zhang
Yu Zhang

18 March 2020 in Optics Letters
Verified
Journal Article
Closed access
Nan WangNan Wang
Nan Wang
WJWei Jiang
Wei Jiang
YZYu Zhang
Yu Zhang
Our group proposes an improved misalignment measurement scheme using the moiré beat signal. Compared with the coarse-fine moiré-based alignment methods, this scheme could complete the nanometer-scale alignment within a centimeter-scale scope in one step. Moreover, it could also fundamentally eliminate the influence from the field of view of the observation lens. These merits make it suitable for the high-precision large-scope misalignment sensing in the proximity, x-ray, and nanoimprint lithographies. The experimental results are given to verify the feasibility and rationality.

25 August 2022 in Optics Letters
Verified
Journal Article
Closed access
PDF file
Nan WangNan Wang
Nan Wang
YLYi Li
Yi Li
FSFeng Sha
Feng Sha
YHYu He
Yu He
Lithography for the next generation of integrated-circuit manufacturing at the 3 nm node requires sub-1-nm misalignment measurement accuracy, which is almost impossible for existing systems due to the optical diffraction limit. Herein, we propose a misalignment sensing strategy based on structured illumination. By virtue of the distinctive modulation effect of a Talbot diffractive illuminated field on moiré fringes, the measurement signals can pass unhindered through the optical system and be used for sensing. Experiments are used to demonstrate that the proposed method can implement real-time-lapse (100 Hz) misalignment sensing with an accuracy of sub-1-nm (0.31 nm @ 3σ), making it suitable for various lithography techniques (e.g., proximity, x ray, projective, and nanoimprint lithography) and fields requiring advanced precision measurement (e.g., quantum measurement, gravitational wave detection, and molecular biology).