Enhancing the metrological performance of non-raster scanning probe microscopy using Gaussian process regression

Zhou, Weiyan; Ren, Mengye; Tao, Yufei; Sun, Lijian; Zhu, Limin

doi:10.1088/1361-6501/ab1d27

Cited by 12 publications

(13 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Accordingly, we present two methods for reconstructing sparse PFM spiral scans: 1) compressive sensing [ 54–60 ] (CS) and 2) Gaussian process [ 61–65 ] (GP) regression. The compressive sensing image inpainting algorithm requires two inputs: 1) The sparse and noisy measurements y , and 2) the scanning mask indicating sampled pixel locations.…”

Section: Introductionmentioning

confidence: 99%

Fast Scanning Probe Microscopy via Machine Learning: Non‐Rectangular Scans with Compressed Sensing and Gaussian Process Optimization

Kelley

Ziatdinov

Collins

et al. 2020

Small

View full text Add to dashboard Cite

Fast scanning probe microscopy enabled via machine learning allows for a broad range of nanoscale, temporally resolved physics to be uncovered. However, such examples for functional imaging are few in number. Here, using piezoresponse force microscopy (PFM) as a model application, a factor of 5.8 reduction in data collection using a combination of sparse spiral scanning with compressive sensing and Gaussian process regression reconstruction is demonstrated. It is found that even extremely sparse spiral scans offer strong reconstructions with less than 6% error for Gaussian process regression reconstructions. Further, the error associated with each reconstructive technique per reconstruction iteration is analyzed, finding the error is similar past ≈15 iterations, while at initial iterations Gaussian process regression outperforms compressive sensing. This study highlights the capabilities of reconstruction techniques when applied to sparse data, particularly sparse spiral PFM scans, with broad applications in scanning probe and electron microscopies.

show abstract

Section: Introductionmentioning

confidence: 99%

Fast Scanning Probe Microscopy via Machine Learning: Non‐Rectangular Scans with Compressed Sensing and Gaussian Process Optimization

Kelley

Ziatdinov

Collins

et al. 2020

Small

View full text Add to dashboard Cite

show abstract

“…Dependent Gaussian process (DGP) [25][26][27] is an extension of the traditional Gaussian process (GP) [23,24]. GP is a well-known non-parametric Bayesian inference method and it captures various features of a specimen through a simple parameterization.…”

Section: Dependent Gaussian Process In Spmsmentioning

confidence: 99%

“…This paper presents a dependent Gaussian process (DGP) regression model for multi-frame semi-sparse image reconstruction, which is an extension of the ordinary Gaussian process (GP) regression model for single image reconstruction [23,24]. A dynamic process is scanned through the semi-sparse scanning pattern, such as sparse raster scanning pattern and sinusoidal scanning pattern.…”

Section: Introductionmentioning

confidence: 99%

Reconstruction of multi-frame semi-sparse scanning probe microscopy images using dependent Gaussian process

Zhou

Ren

Zhu

2020

Meas. Sci. Technol.

View full text Add to dashboard Cite

A major drawback of scanning probe microscopes (SPMs) is the time-consuming image acquisition, typically on the order of tens of seconds to minutes. Sparse raster scanning and non-raster scanning have been demonstrated to be effective approaches to improve the efficiency of SPMs. However, these strategies entail the challenge of reconstructing accurate images from semi-sparse sampling points, which means the distribution of the scattered sampling points is considerably sparser in the direction perpendicular to the scanning path. In this paper, we focus on the multi-frame semi-sparse video image reconstruction. In this setting, one of the frames is regarded as the reference frame while the others are regarded as the target frames. Instead of reconstructing the target frames independently, we propose a dependent Gaussian process regression model to make full use of the information provided by the reference frame and the currently processed target frame. The proposed method takes advantage of the intrinsic correlation between images. Thus, the accuracy of the reconstruction can be improved greatly. Comparisons with the renowned Delaunay triangulation-based interpolation method and the Gaussian process regression method prove that the proposed method has dramatically improved the reconstruction accuracy of semi-sparse video images.

show abstract

“…However, the broad adoption of sparse scanning necessitates the development of algorithmic tools to achieve full information recovery. This is now an option thanks to the advent of machine learning image reconstruction algorithms such as compressed sensing (CS) 29 , convolutional neural networks 34 (CNN) or Gaussian process (GP) optimization 35,36 , which employ sparse data schemes with subsequent reconstruction of non-imaged areas. Combining sparse, nonrectangular scans, with such algorithms offer several advantages to SPM, including minimized perturbation to the system, extended probe life, and importantly their combination presents clear opportunities for faster image acquisition.…”

Section: Introductionmentioning

confidence: 99%

High speed mapping of surface charge dynamics via Spiral Scanning Kelvin Probe Force Microscopy

Checa

Kelley

Sun

et al. 2023

Preprint

View full text Add to dashboard Cite

Understanding local dynamic charge processes is essential for developing advanced materials and devices, from batteries and microelectronics to medicine and biology. Continued progress relies on the ability to map electronic and ionic transport phenomena across different time and length scales, encompassing the intrinsic heterogeneities of the material itself (e.g., grain boundaries, domain walls, etc.). To address this challenge, we introduce high-speed Spiral-Scanning Kelvin Probe Force Microscopy (SS-KPFM), which combines sparse spiral scanning and image reconstruction via Gaussian process optimization. SS-KPFM enables functional sub-second imaging rates (≈ 3 fps), which represents a significant improvement over current state-of-the-art and several orders of magnitude over traditional KPFM methods. We apply it to study the spatiotemporal charge dynamics at a LaAlO3/SrTiO3 planar device and charge injection and diffusion dynamics in polycrystalline TiO2 thin films, providing full 2D Contact Potential Difference (CPD) maps of the surface charge dynamics in a fast and automated fashion.

show abstract

Enhancing the metrological performance of non-raster scanning probe microscopy using Gaussian process regression

Cited by 12 publications

References 32 publications

Fast Scanning Probe Microscopy via Machine Learning: Non‐Rectangular Scans with Compressed Sensing and Gaussian Process Optimization

Fast Scanning Probe Microscopy via Machine Learning: Non‐Rectangular Scans with Compressed Sensing and Gaussian Process Optimization

Reconstruction of multi-frame semi-sparse scanning probe microscopy images using dependent Gaussian process

High speed mapping of surface charge dynamics via Spiral Scanning Kelvin Probe Force Microscopy

Contact Info

Product

Resources

About