Mircea Cimpoi

Mircea Cimpoi

Research Engineer combining first-author publications, production-grade ML infrastructure, and ethical data collection experience.

  • Creator of the Describable Textures Dataset (DTD; CVPR 2014; 4,500+ citations and counting).
  • Previously at Meta and Sony AI; earlier at Microsoft.
  • PhD from the Oxford, VGG (2015) and postdoc at CIIRC, CTU in Prague.

Google Scholar | GitHub | LinkedIn | Resume | Thesis | Conference Deadlines

I run experiments for work. I run half marathons for fun. Always run the extra mile.

Publications

Complete list on Google Scholar.
img
Neighbourhood Consensus Networks
I. Rocco, M. Cimpoi, R. Arandjelović, A. Torii, T. Pajdla, J. Sivic
In 32nd Neural Information Processing Systems (NeurIPS), 2018. Spotlight
Project Page | Paper | Code and Data | Spotlight Video
img
InLoc: Indoor Visual Localization with Dense Matching and View Synthesis
H. Taira, M. Okutomi, T. Sattler, M. Cimpoi, M. Pollefeys, J. Sivic, T. Pajdla, A. Torii
In Computer Vision and Pattern Recognition (CVPR), 2018. Spotlight
Project Page | Paper | arXiv | Code and Data
deep-fb
Deep Filter Banks for Texture Recognition, Description and Segmentation
M. Cimpoi, S. Maji, I. Kokkinos, A. Vedaldi
In International Journal of Computer Vision (IJCV), 2016,
Project Page | Paper | Code
deep-fb
Deep Filter Banks for Texture Recognition and Segmentation
M. Cimpoi, S. Maji, A. Vedaldi
In Computer Vision and Pattern Recognition (CVPR), 2015. Oral
Project Page | Paper | Code | Data
deep-fb
Describing Textures in the Wild
M. Cimpoi, S. Maji, I. Kokkinos, S. Mohamed, A. Vedaldi
In Computer Vision and Pattern Recognition (CVPR), 2014.
Project Page | Paper | Code | Data: Website | Tensorflow | torchvision

Reviewing

CVPR: 2024; NeurIPS: 2019 - 2022, 2024 (Top Reviewer), TPAMI: 2021, 2018; ICLR: 2021.

Initial version inspired and adapted from Jon Barron's website.