Machine Learning Computer Vision

Guangyu Hu

Ph.D. candidate, University of Arizona

Graduate research assistant in materials science and machine learning Phone: 520-***-**** Email: ***@*******.*** Github: github.com/hugy888 Professional Summary

Multidisciplinary research professional with over 4 years of hands-on experience in semiconductor processing, thin-film fabrication, and material characterization. Proficient in a range of industry-standard fabrication and analysis techniques, including photolithography, E-beam deposition, and wafer handling. Additionally, I have 3 years of experience in machine learning and data analytics, leveraging these skills to facilitate the design, manufacturing, and characterization of materials. Skills

- Semiconductor Processing: Photolithography, Wet Etching, Wafer Handling, E-beam Deposition, Thermal Oxidation, Cleanroom Protocols

- Thin Film Fabrication: X-ray Reflectivity (XRR), Atomic Force Microscopy (AFM), Transmission Electron Microscopy (TEM), UV-Vis Spectroscopy

- Machine Learning & Data Analysis: Graph Neural Networks, Computer Vision (Segment Anything, DINOv2), Statistical Multi-variable Data Analysis

- Programming: Python (PyTorch, TensorFlow), MATLAB, Java, C/C++

- Scientific Software: Solidworks, AutoCAD, COMSOL, Thermo-Calc Internship Experience

Materials Compatibility Intern (Manager: Rahul Patil) 2024 Summer Roche Tissue Diagnostics, Tucson, AZ

- Automated evaluation of cytology image quality using computer vision techniques, streamlining diagnostic processes.

- Developed high-throughput image segmentation tools for cytology slide analysis on HPC clusters.

Mechanical Engineer Intern (Supervisor: Shanhui Xu) 2015 Spring South China University of Technology, Guangzhou, China

- Designed machine elements for laser devices using Solidworks and AutoCAD.

- Conducted testing and debugging of laser machines in a cleanroom environment, adhering to precise quality standards.

Research Experience

University of Arizona (Advisor: Zafer Mutlu) 2024 Spring

- Fabricated metal-oxide-semiconductor capacitors in a cleanroom setting at the UA Nano Fab Center.

- Conducted thin-film deposition and characterization using state-of-the-art techniques including E-beam deposition and photolithography.

University of Arizona (PhD Advisor: Marat Latypov) 2021-Present

- Developed machine learning models for optimizing microstructure-property relationships in materials, enhancing semiconductor device performance.

- Fine-tuned computer vision models for automated segmentation of melt-pools of 3D printed metals.

- Utilized thermodynamic dislocation theory to model stress-strain behavior in semiconductor materials.

Rensselaer Polytechnic Institute (MSc Advisor: Peter Dinolfo) 2017-2018

- Fabrication & characterization of organic thin films and self-assembled monolayers (SAM).

- Studied photophysical properties of thin films using UV-Vis absorption and fluorescence spectroscopy.

- Conducted thin film thickness measurement using X-ray Reflectivity (XRR) and in-plane GIXRD for SAM analysis.

South China University of Technology (BSc Advisor: Xiaodong Cao) 2014-2016

- Prepared and optimized nanoliposomes using film dispersed-ultrasonic technology.

- Proved the encapsulation of Vitamin E inside liposomes using IR, DSC, and SEM. Publications

1. G. Hu, M.I. Latypov. (under preparation) Graph neural networks for predicting properties of a wide class of polycrystals.

2. G. Hu, M.I. Latypov. AnisoGNN: Graph neural networks generalizing to anisotropic properties of polycrystals. Computational Materials Science, 243, (2024): 113121. 3. S.E. Whitman, G. Hu, M.I. Latypov. Learning microstructure-property relationships in materials with robust features from vision transformers. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, (2024): 8125-8130. 4. S.E. Whitman, G. Hu, H.C. Taylor, R.B. Wicker, M.I. Latypov. Automated segmentation and chord length distribution of melt pools in complex 3D printed metal artifacts. Integrating Materials and Manufacturing Innovation, (2023): 1-15. 5. G. Hu, M.I. Latypov. Aniso-GNN: physics-informed graph neural networks that generalize to anisotropic properties of polycrystals. NeurIPS 2023 Workshop AI4Mat. 6. G. Hu, M.I. Latypov. Learning from 2D: machine learning of 3D effective properties of heterogeneous materials based on 2D microstructure sections. Frontiers in Metals and Alloys, 1 (2022): 1100571.

7. G. Hu, Z. Huang, P. Dinolfo. (under preparation) PDI Molecular Structure Induced Aggregation Effects in Multilayer Thin Films on Optical Properties. 8. G. Hu. Molecular Structure Induced Aggregation Effects on the Photophysical Properties of Perylene Diimide Based Multilayer Thin Films. 2019. 9. X. Cao, Y. Zhang, G. Hu, Z. Wu, X. Zhao, Y. Chen, Y. Huang, S. Xiao. Preparation Method and Application of Polyethylene Glycol - Modified Vitamin E Liposomes. CN Patent 105055185, 2018.

10. G. Hu, Z. Wu, Y. Chen, Y. Zhang, X. Zhao, S. Xiao, X. Cao. Preparation and Property of Polyethylene Glycol Modified VE - Containing Flexible Liposomes. Flavour Fragrance Cosmetics, 6 (2016): 44-48.

Contact this candidate