I work on machine-learning code where the data is visual, spatial, or clinical.

At NYU Langone Health, I build research software around healthcare problems. I am also pursuing an M.S. in Computer Science through Georgia Tech.

Before that: machine-learning engineering at Woven by Toyota, GeoAI tooling at Esri, and research visits with Yuke Zhu and Mohamed Elhoseiny.

Portrait of Divyansh Jha

About

Focus Computer vision, robotics, geospatial deep learning, and healthcare research software.
Pattern Mostly model training, evaluation, data preparation, and the glue code that keeps research work reproducible.
ArcGIS Worked on the learn module of ArcGIS API for Python: data preparation, model training, and deep-learning workflows for GIS users.
Contact divyanshj.16@gmail.com GitHub Twitter LinkedIn

Work

Period Role Organization
Aug 2024 - Present Research Engineer NYU Langone Health / healthcare research software
Dec 2021 - Jul 2024 Machine Learning Engineer Woven by Toyota / machine-learning engineering
Jan 2021 - Oct 2021 Visiting Researcher The University of Texas at Austin, with Dr. Yuke Zhu
Feb 2020 - Aug 2021 Research Intern King Abdullah University of Science and Technology, with Dr. Mohamed Elhoseiny
May 2019 - Nov 2021 Data Scientist II Esri R&D Center / deep learning for ArcGIS
May 2018 - May 2019 Machine Learning Intern Esri R&D Center / early GeoAI work

Education

Years Program Institution
2024 - 2027 M.S. Computer Science Georgia Institute of Technology
2015 - 2019 B.Tech Electronics and Communication Engineering Maharaja Agrasen Institute of Technology, Guru Gobind Singh Indraprastha University

Writing

Topic Post Notes
Generative modeling Implementing SPADE using fastai A hands-on implementation note for semantic image synthesis.
GeoAI Swimming pool detection and classification using deep learning Detection and classification work on geospatial imagery.
Computer vision Not just another GAN paper - SAGAN Notes on self-attention in generative adversarial networks.
Robustness Tackling Adversarial Examples : Introspective CNN A short read on adversarial examples and image classifiers.