Stanislav Pidhorskyi
Graduate Research Assistant
Morgantown, West Virginia, United States
Summary
🤩
Rockstar🎓
Top SchoolStanislav Pidhorskyi is a PhD student and graduate research assistant at West Virginia University specializing in computer vision, generative models, and novelty/anomaly detection, with 11 years of engineering and research experience. He transitioned from a production-focused game development career—where he implemented advanced graphics techniques like LiSPSM shadow mapping and text/shaping support using HarfBuzz at Gameloft—into machine learning research, blending low-level graphics expertise with probabilistic modeling. Stanislav has hands-on ML engineering experience contributing to public generative model code (e.g., improvements to ALAE's VAE and style-based layers) and interned at AWS Rekognition, signaling applied industry familiarity with vision systems. His background in aerospace mechanics and long-standing software development experience give him a strong foundation in numerical methods and performance-conscious implementation.
11 years of coding experience
6 years of employment as a software developer
PhD, Computer Science, PhD, Computer Science at West Virginia UniversityMaster's Degree, Dynamics and Strength of Machines, Master's Degree, Dynamics and Strength of Machines at National Aerospace University -'Kharkiv Aviation Institute'
Stackoverflow
1,572reputation
184kreached
32answers
4questions
opengl
top-5%
Github Skills (15)
pytorch10
machine-learning10
image-generation10
autoencoder10
opengl9
alpha-blending6
opengl-es6
projection6
rotation6
matrix6
shader6
glsl6
3d6
framebuffer6
tensorflow3
Programming languages (8)
Github contributions (5)
podgorskiy/ALAE
Mar 2019 - Dec 2020
[CVPR2020] Adversarial Latent Autoencoders
Role in this project:
ML Engineer
ML EngineerContributions:292 commits, 5 PRs, 21 pushes in 1 year 9 months
Contributions summary:Stanislav's commits primarily focus on modifications to the VAE (Variational Autoencoder) and related files, particularly `VAE.py` and `net.py`, suggesting a focus on implementing and refining the autoencoder architecture. The changes include modifications to the loss function, the introduction of style-based layers, and integration with a discriminator. This indicates the user is actively involved in improving and experimenting with the autoencoder, possibly for image generation or related tasks.
pytorchautoencoderunsupervised-learningdeep-learningadversarial
podgorskiy/PicoRender
Aug 2015 - Oct 2020
Contributions:18 pushes, 4 branches, 4 tags in 5 years 3 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.