Skip to content
View nikitadurasov's full-sized avatar
🥷
🥷

Highlights

  • Pro

Block or report nikitadurasov

Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
nikitadurasov/README.md

Hello! 🌱

I’m Nikita Durasov, a Research Scientist at NVIDIA, bridging the gap between state-of-the-art research and scalable AV products by building robust perception and planning for self-driving cars.

GitHub followers Visits Badge Google Scholar Website LinkedIn Twitter

  • 🔭 Research: Published in CVPR, ICCV, ICML on topics like neural network robustness, uncertainty, and computer vision.
  • 🌱 Industry Experience: Collaborated with NVIDIA, Apple, Amazon, and Samsung on AI and ML solutions.
  • 🎓 Education: PhD in Computer Science from EPFL, where I focused on model uncertainty and generalization.

🛠️ Tools & Technologies

Python PyTorch TensorFlow scikit-learn Lightning OpenCV HuggingFace

Pinned Loading

  1. masksembles masksembles Public

    Official repository for the paper "Masksembles for Uncertainty Estimation" (CVPR 2021).

    Python 103 16

  2. torch-ttt torch-ttt Public

    A modular and easy-to-use framework for Test-Time Training (TTT) and Test-Time Adaptation (TTA) in Pytorch, making your networks more generalizable with minimal effort ✨

    Python 27 2

  3. cvlab-epfl/iter_unc cvlab-epfl/iter_unc Public

    Official code for "Enabling Uncertainty Estimation in Iterative Neural Networks" (ICML 2024)

    Jupyter Notebook 18 1

  4. cvlab-epfl/zigzag cvlab-epfl/zigzag Public

    Official code for "ZigZag: Universal Sampling-free Uncertainty Estimation Through Two-Step Inference" (TMLR 2024)

    Jupyter Notebook 17 4

  5. ittt ittt Public

    Official code for "IT³: Idempotent Test-Time Training" (ICML 2025)

    Jupyter Notebook 15 1

  6. SPD_reduce SPD_reduce Public

    This repo is dedicated to Deep Manifold Learning of Symmetric Positive Definite (SPD) Matrices with Application to Face Recognition article from AAAI 2017.

    Python 7 3