Jakob Drachmann Havtorn*, *, Tijmen Blankevoort, Babak Ehteshami Bejnordi
in ICCV Workshop on New Ideas in Vision Transformers (NViT), 2023
Resume
Work
| 2021 - today | Deep Learning Research Engineer Qualcomm AI Research 🇳🇱 Amsterdam, The Netherlands |
|---|---|
Education
| 2015 - 2020 | PhD student at ISTA Institute of Science and Technology Austria, Klosterneuburg, Austria |
|---|---|
| 2012 - 2015 | Masters and Bachelors in Computer science at Ecole Normale Superieure (ENS) de Rennes In conjunction with studies at Irisa / University of Rennes 1, Rennes, France |
| 2010 - 2012 | Classes Préparatoires aux Grandes Écoles (CPGE) Lycée Georges Clémenceau, Reims, France, MPSI-MP * |
Internships
| 2020 |  | Research internship at Google Brain Zürich 🇨🇭 5 months, Zürich, Switzerland Topic: Working on large neural networks compression via knowledge distillation, advised by Lucas Beyer and Alexander Kolesnikov |
|---|---|---|
| 2017 |  | Research internship at Google Brain London 🇬🇧 London, United Kingdom Topic: Unsupervised image-to-image translation combining GANs and domain adaptation techniques, advised by Konstantinos Bousmalis, Stephan Gouws and Fred Bertsch |
| 2015 | | Internship at Inria Rennes 🇫🇷 Rennes, France Topic: Unsupervised clustering via random classifiers for text and audio clustering, advised by Vincent Claveau and Guillaume Gravier |
| 2014 |  | Internship at IST Austria 🇦🇹 Vienna, Austria Topic: Adapting pre-trained classifiers to unknown test labels distribution on-the-fly, advised by Christoph Lampert |
| 2013 |  | Internship at Inria Rennes 🇫🇷 Rennes, France Topic: Video retrieval using circular Fourier transforms, advised by Hervé Jégou and Teddy Furon |
Selected Publications
- MSViT: Dynamic Mixed-Scale Tokenization for Vision Transformers
- Scalarization for Multi-Task and Multi-Domain Learning at Scale
, Tijmen Blankevoort, Babak Ehteshami Bejnordi
in Conference on Neural Information Processing Systems (NeurIPS), 2023 - Knowledge Distillation: A good teacher is patient and consistent
Lucas Beyer, Xiaohua Zhai, , Larisa Markeeva, Rohan Anil, Alexander Kolesnikov
in Conference on Computer Vision and Pattern Recognition (CVPR) (oral), 2022 - Localizing Grouped Instances for Efficient Detection in Low-Resource Scenarios
and Christoph Lampert
in Winter Conference on Applications of Computer Vision (WACV), 2020 - XGAN: Unsupervised Image-to-Image Translation for Many-to-Many Mappings
, Konstantinos Bousmalis, Stephan Gouws, Fred Bertsch, Inbar Mosseri, Forrester Cole, Kevin Murphy
in Domain Adaptation for Visual Understanding Workshop at ICML/IJCAI/EJCAI 2018, 2018 - Probabilistic Image Colorization
, and Christoph Lampert
in British Machine Vision Conference (BMVC), 2017 - Classifier Adaptation at Prediction Time
and Christoph Lampert
in Conference on Computer Vision and Pattern Recognition (CVPR), 2015
Skills
- Languages
- 🇫🇷 French (native speaker)
- 🇬🇧 English (fluent)
FCE/B2 in 2010, TOEIC (990/990) in 2013 - 🇩🇪 German (advanced)
ZD/B1 in 2008, ZMP/C1 in 2010
- Software
Python,C++/C,OCaml- Deep Learning:
Pytorch,Tensorflow,Jax,Keras - Visualization:
matplotlib,streamlit,LateX - Version control:
git