Hello
My name is Vitaliy, I specialize in computer vision with a focus on generative neural networks (GAN). I have been working on projects with GAN technology for the last 2 years.
In general, I am fond of scientific programming; I study interesting algorithms and articles, as well as software architecture. I like fast code, I can optimize both the algorithm and the implementation, including at a low level.
I am able to solve non-standard tasks, I work for the result and responsibly :)Projects I've done at different times:
- (2022) PyTorch Implementation of "Progressive Distillation for Fast Sampling of Diffusion Models(v-diffusion)". github
- (2020–2021) Virtual trainer computer vision app for Android: counts the number of repetitions of the exercise, with my own neural nets. Work in realtime (~8 fps). Used technologies: pytorch, opencv, c++, android sdk/ndk. youtube
- (2020) Modified version of StarGANv2 with increased resolution 512x512. medium
- (2011) Experimental sensetive data protection system against attacks through vulnerabilities in users software for Linux workstations github (C++/Qt)
- (2010) A system for automating the search of logical errors in the program code. II-nd place in the SIBINFO information security competition. presentation (in Russian)
- (2007) The experimental 3D reconstruction system. I-st place at the Ural Interregional Conference. medium