I am a PhD student at Computer Vision and Geometry Group, ETH Zurich under the supervision of Prof. Marc Pollefeys and Dr. Lubor Ladicky. Before joining Computer Vision and Geometry Group, I worked at the Russian search engine Yandex for 2 years.
My goal is to give machines the most complete and accurate representation of the real 3D world and help to manipulate it. I've worked on Semantic 3D Reconstruction, shape priors and invariances in deep learning. Currently I'm interested in unsupervised and reinforcement (deep) learning.
Visit my google scholar profile.
- N. Savinov, A. Seki, L. Ladicky, T. Sattler, M. Pollefeys, Quad-networks: unsupervised learning to rank for interest point detection
- N. Savinov, C. Häne, L. Ladicky, M. Pollefeys, Semantic 3D Reconstruction with Continuous Regularization and Ray Potentials Using a Visibility Consistency Constraint, CVPR 2016 (spotlight oral, acceptance rate 9.7%)
[pdf] [spotlight talk] [spotlight slides] [poster] [longer talk] [longer talk slides]
- D. Laptev, N. Savinov, J. Buhmann, M. Pollefeys, TI-POOLING: transformation-invariant pooling for feature learning in Convolutional Neural Networks, CVPR 2016 (acceptance rate 30%)
- N. Savinov, L. Ladicky, C. Häne, M. Pollefeys, Discrete Optimization of Ray Potentials for Semantic 3D Reconstruction, CVPR 2015 (acceptance rate 28%)
[pdf] [video] [poster]
- C. Häne, N. Savinov, M. Pollefeys, Class Specific 3D Object Shape Priors Using Surface Normals, CVPR 2014 (acceptance rate 30%)
nikolay dot savinov at inf dot ethz dot ch
+41 (0)44 63 33 714
CAB G 81.1
Computer Vision and Geometry Group
Department of Computer Science