Mikita Sazanovich

Won the Lane Following competition of the AI-DO 3 at NeurIPS 2019! I implemented an imitation learning algorithm, developed its training regime on several datasets and built its architecture with the independence of every neuron in mind. The code is available on GitHub.

[December 2019]

Trained a Dota 2 agent using replays of high-rated players available online. Check out the code on GitHub. You can also watch the sample episodes of the agent here and here.

[June 2019]

Participated in the AI-DO 2 competition at ICRA 2019. We won the second edition of the competition by integrating a road segmentation algorithm into a conventional autonomous driving stack. The code for the method is on GitHub.

[May 2019]

Placed 4-th in the first edition of the AI Driving Olympics (AI-DO 1) by using an end-to-end reinforcement learning algorithm. You can find the code on GitHub.

[December 2018]

Developed a plugin for IntelliJ based IDEs that allows you to lookup relevant to your editor context documentation on Stack Overflow. The plugin is available for download and open-sourced.

[June 2018]

Placed at the top 10%, winning a silver medal at the International Olympiad in Informatics (IOI) 2015.

[August 2015]