Deep-Learning-Based Image Segmentation Integrated with Optical Microscopy for Automatically Searching for Two-Dimensional Materials

Abstract

Deep-learning algorithms enable precise image recognition based on high-dimensional hierarchical image features. Here, we report the development and implementation of a deep-learning-based image segmentation algorithm in an autonomous robotic system to search for two-dimensional (2D) materials. We trained the neural network based on Mask-RCNN on annotated optical microscope images of 2D materials (graphene, hBN, MoS2, and WTe2). The inference algorithm is run on a 1024 x 1024 px2 optical microscope images for 200 ms, enabling the real-time detection of 2D materials. The detection process is robust against changes in the microscopy conditions, such as illumination and color balance, which obviates the parameter-tuning process required for conventional rule-based detection algorithms. Integrating the algorithm with a motorized optical microscope enables the automated searching and cataloging of 2D materials. This development will allow researchers to utilize unlimited amounts of 2D materials simply by exfoliating and running the automated searching process.

References in corpus (14)

• Electric Field Effect in Atomically Thin Carbon Films
• SSD: Single Shot MultiBox Detector
• Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
• A Survey on Deep Learning in Medical Image Analysis
• Experimental Observation of Quantum Hall Effect and Berry's Phase in Graphene
• Two Dimensional Atomic Crystals
• TensorFlow: A system for large-scale machine learning
• 2D materials and van der Waals heterostructures
• A Review on Deep Learning Techniques Applied to Semantic Segmentation
• Rapid and reliable thickness identification of two-dimensional nanosheets using optical microscopy
• Deep Learning Scaling is Predictable, Empirically
• Intelligent Identification of Two-Dimensional Structure by Machine-Learning Optical Microscopy
• Thickness determination of MoS2, MoSe2, WS2 and WSe2 on transparent stamps used for deterministic transfer of 2D materials
• A Survey on Deep Learning of Small Sample in Biomedical Image Analysis

Cited by in corpus (16)

• Ultra-clean assembly of van der Waals heterostructures
• Benchmark and application of unsupervised classification approaches for univariate data
• Universal image segmentation for optical identification of 2D materials
• Quantum Optics Applications of Hexagonal Boron Nitride Defects
• An open-source robust machine learning platform for real-time detection and classification of 2D material flakes
• Fully automated identification of 2D material samples
• Phyllosilicates as earth-abundant layered materials for electronics and optoelectronics: Prospects and challenges in their ultrathin limit
• Identification and Structural Characterization of Twisted Atomically Thin Bilayer Materials by Deep Learning
• Predicting phase preferences of two-dimensional transition metal dichalcogenides using machine learning
• Selective Etching of Hexagonal Boron Nitride by High-Pressure CF4 Plasma for Individual One-dimensional Ohmic Contacts to Graphene Layers
• Improving Classification through Weak Supervision in Context-specific Conversational Agent Development for Teacher Education
• Deep-learning Assisted Detection and Quantification of (oo)cysts of Giardia and Cryptosporidium on Smartphone Microscopy Images
• MaskTerial: A Foundation Model for Automated 2D Material Flake Detection
• Machine Learning-based Automatic Graphene Detection with Color Correction for Optical Microscope Images
• Direct observation of the layer-number-dependent electronic structure in few-layer WTe2
• Physics-informed neural networks for solving thermo-mechanics problems of functionally graded material