Abstract:The lack of available annotated Scanning Probe Microscopy (SPM) data limits the applicability of Machine Learning to evaluate images on the nanoscale, requiring laborious manual work. The use of synthetic data to train state‐of‐the‐art object detection models can overcome this limitation, enabling a fast, accurate, and reliable analysis of SPM images, as demonstrated on two self‐assembled molecular chiral networks. Enantiospecific effects play an uprising role in chemistry and technical applications. Chiral molecular networks formed by self‐assembly processes at surfaces can be imaged by scanning probe microscopy (SPM). Low contrast and high noise in the topography map often interfere with the automatic image analysis using classical methods. The long SPM image acquisition times restrain Artificial Intelligence‐based methods requiring large training sets, leaving only tedious manual work, inducing human‐dependent errors and biased labeling. By generating realistic looking synthetic images, the acquisition of real datasets is avoided. Two state‐of‐the‐art object detection architectures are trained to localize and classify chiral unit‐cells in a regular molecular chiral network formed by self‐assembly of linear molecular bricks. The comparison of different architectures and datasets demonstrates that the training on purely synthetic data outperforms models trained using augmented datasets. A Faster R‐CNN model trained solely on synthetic data achieved an excellent mean average precision of 99% on real data. Hence this approach and the transfer to real data show high success, also highlighting the high robustness against experimental noise and different zoom levels across the full experimentally reasonable parameter range. The generalizability of this idea is demonstrated by achieving equally high performance on a different structure, too.

Testing different supervised machine learning architectures for the classification of liquid crystals

Classification of liquid crystal textures using convolutional neural networks

Automatic Phase Recognition Method Based on Convolutional Neural Network

Deep learning techniques for the localization and classification of liquid crystal phase transitions

Learning physical properties of liquid crystals with deep convolutional neural networks

Convolutional Neural Network analysis of optical texture patterns in liquid-crystal skyrmions

Determining liquid crystal properties with ordinal networks and machine learning

Classification of crystal structures using electron diffraction patterns with a deep convolutional neural network

Optical Convolutional Neural Network Based on an Amplitude Modulation Spatial Light Modulator and a 4-Level Phase Plate

A Siamese neural network framework for glass transition recognition.

Prediction of the Structural Color of Liquid Crystals via Machine Learning

Performance of a deep convolutional neural network to classify crystal structures using selected area electron beam diffraction patterns containing lattice defect information

Chirality Detection in Scanning Tunneling Microscopy Data Using Artificial Intelligence

Classification of Astronomical Bodies by Efficient Layer Fine-Tuning of Deep Neural Networks

Refining Architectures of Deep Convolutional Neural Networks

Machine learning methods for liquid crystal research: phases, textures, defects and physical properties

Transitioning between Convolutional and Fully Connected Layers in Neural Networks

Optical characterization of dyed liquid crystal cells

Intra-class Classification of Architectural Styles Using Visualization of CNN

Inception Architecture and Residual Connections in Classification of Breast Cancer Histology Images