Nikita Dvornik

PhD in Machine Learning and Computer Vision

I am a research scientist at Samsung AI Center Toronto, working on video understanding.

Previously, I was a postdoc at UofT working on unsupervised learning from visual observations, research intern at Uber ATG working on autonomous driving, PhD at Inria Grenoble working on visual understanding and few-shot learning, MSc in computer science at INP Grenoble, and BSc in applied math at MIPT.

Google scholar   GitHub   Twitter   LinkedIn   Email   CV  

  1. arXiv’22
    SlotFormer: Unsupervised Visual Dynamics Simulation with Object-Centric Models
    Ziyi Wu, Nikita Dvornik, Klaus Greff, Thomas Kipf, and Animesh Garg
    arXiv preprint arXiv:2210.05861 2022
    arXiv HTML
  2. ECCV’22
    Oral
    Flow graph to Video Grounding for Weakly-supervised Multi-Step Localization
    Nikita Dvornik, Isma Hadji, Hai Pham, Dhaivat Bhatt, Brais Martinez, Afsaneh Fazly, and Allan D Jepson
    In European Conference on Computer Vision (ECCV) 2022
    arXiv Code
  3. CVPR’22
    Oral
    P3IV: Probabilistic Procedure Planning from Instructional Videos with Weak Supervision
    He Zhao, Isma Hadji, Nikita Dvornik, Konstantinos G Derpanis, Richard P Wildes, and Allan D Jepson
    In Conference on Computer Vision and Pattern Recognition (CVPR) 2022
    arXiv Code
  4. NeurIPS’21
    Drop-DTW: Aligning Common Signal Between Sequences While Dropping Outliers
    Nikita Dvornik, Isma Hadji, Konstantinos G Derpanis, Animesh Garg, and Allan D Jepson
    In Advances in Neural Information Processing Systems (NeurIPS) 2021
    arXiv Code
  5. ECCV’20
    Selecting relevant features from a multi-domain representation for few-shot classification
    Nikita Dvornik, Cordelia Schmid, and Julien Mairal
    In European Conference on Computer Vision (ECCV) 2020
    arXiv Code
  6. ICCV’19
    Diversity with cooperation: Ensemble methods for few-shot classification
    Nikita Dvornik, Cordelia Schmid, and Julien Mairal
    In International Conference on Computer Vision (ICCV) 2019
    arXiv Code
  7. ECCV’18
    Modeling visual context is key to augmenting object detection datasets
    Nikita Dvornik, Julien Mairal, and Cordelia Schmid
    In European Conference on Computer Vision (ECCV) 2018
    arXiv Code
  8. ICCV’17
    Blitznet: A real-time deep network for scene understanding
    Nikita Dvornik, Konstantin Shmelkov, Julien Mairal, and Cordelia Schmid
    In International Conference on Computer Vision (ICCV) 2017
    arXiv Code