I am postdoctoral researcher at King Abdullah University of Science and Technology (KAUST) working with Prof. Bernard Ghanem. I completed my Phd at the University of Amsterdam, advised by Prof. Cees Snoek. My area of interest is Video Understanding, with my PhD thesis (which you can find here ) focussing on Video-Efficient Foundation Models. I am particularly interested in training foundation models via self-supervised learning from multiple modalities of the video data.

Contact: f.m.thoker *at* uva.nl

News & Activities


Publications


SIGMA: Sinkhorn-Guided Masked Video Modeling
Mohammadreza Salehi*, Michael Dorkenwald*, Fida Mohammad Thoker*, Efstratios Gavves, Cees G. M. Snoek, Yuki M. Asano
European Conference on Computer Vision (ECCV), 2024.
[Webpage] [arXiv] [Code]
Tubelet-Contrastive Self-Supervision for Video-Efficient Generalization
Fida Mohammad Thoker, Hazel Doughty, Cees Snoek
International Conference on Computer Vision (ICCV), 2023.
[Webpage] [arXiv] [Code]
How Severe is Benchmark-Sensitivity in Video Self-Supervised Learning?
Fida Mohammad Thoker, Hazel Doughty, Piyush Bagad, Cees Snoek
European Conference on Computer Vision (ECCV), 2022.
[Webpage] [arXiv] [Code]
Skeleton-Contrastive 3D Action Representation Learning
Fida Mohammad Thoker, Hazel Doughty, Cees Snoek
ACM International Conference on Multimedia (ACMMM), 2021
[arXiv] [Code]
Feature-Supervised Action Modality Transfer
Fida Mohammad Thoker, Cees Snoek
IEEE International Conference on Pattern Recognition (ICPR), 2020
[arXiv]
CROSS-MODAL KNOWLEDGE DISTILLATION FOR ACTION RECOGNITION
Fida Mohammad Thoker, Juergen Gall
IEEE International Conference on Image Processing (ICIP), 2019
[arXiv]

Academic Service



Reviewer: BMVC 2020, CVIU 2021, Nuerips 2021, ICCV 2021, ECCV 2022, ACCV 2022, CVPR 2023, ICCV 2023

Teaching



Teaching Assistant: Deep Learning for Visual Recognition (MSc Computer Science Univerisity of Bonn)
Teaching Assistant: Technical Neural Networks (MSc Computer Science Univerisity of Bonn)