Noureldien Hussein

arXiv, 2020

Noureldien Hussein, Efstratios Gavves, Arnold W. M. Smeulders

A novel convolutional block named Permutation Invariant Convolution (PIC) is proposed. It is permissible in many ways that a long-range activity can take place over time. The result is improvements in the accuracy of recognizing the long-range activities of Breakfast, Charades, and MultiThumos.

@inproceedings{hussein2020pic,
title     = {PIC: Permutation Invariant Convolution for Recognizing Long-range Activities},
author    = {Hussein, Noureldien and Gavves, Efstratios and Smeulders, Arnold W.M.},
booktitle = {arXiv},
year      = {2020}
}

arXiv, 2020

Noureldien Hussein, Mihir Jain, Babak Ehteshami Bejnordi

A conditional gating mechanism is proposed to select the most relevant segments from a video of long-range activity. The selection is conditioned on both the segment and its context, i.e. the video itself. The result is extreme efficiency in classifying long-range activities without compromising the performance.

@inproceedings{hussein2020timegate,
title     = {TimeGate: Conditional Gating of Segments in Long-range Activities},
author    = {Hussein, Noureldien and Jain, Mihir and Bejnordi, Babak Ehteshami},
booktitle = {arXiv},
year      = {2020}
}

ICPR, 2020

Mert Kilickaya, Noureldien Hussein, Efstratios Gavves, Arnold Smeulders

For recognizing human-object interaction from single image, we propose new sources of contexts. In addition, we propose a conditional gating model for fusing the context features. The result is a substantial improvement on three benchmarks: HICO, V-COCO and CINT.

@inproceedings{kilickaya2020self,
title     = {Self-Selective Context for Interaction Recognition},
author    = {Kilickaya, Mert; Hussein, Noureldien; Gavves, Efstratios; Smeulders, Arnold W.M.},
booktitle = {ICPR},
year      = {2020}
}

ICCV Workshop, 2019

Noureldien Hussein, Efstratios Gavves, Arnold W. M. Smeulders

A method to represent a long-range action as a graph. A soft version of the graph is reconstructed, where both the graph nodes and graph edges are learned entirely from data. The result is recognizing minutes-long human activities in videos.

@inproceedings{hussein2019videograph,
title     = {VideoGraph: Recognizing Minutes-Long Human Activities in Videos},
author    = {Hussein, Noureldien and Gavves, Efstratios and Smeulders, Arnold W.M.},
booktitle = {ICCV Workshop on Scene Graph Representation and Learning},
year      = {2019}
}

CVPR, 2019 – Oral Presentation

Noureldien Hussein, Efstratios Gavves, Arnold W. M. Smeulders

A novel temporal layer for 3D CNNs with multi-scale temporal convolutions and much reduced computation. The result is a CNN for modeling minute-long complex actions, 8 times longer than best related method.

@inproceedings{hussein2018timeception,
title     = {Timeception for Complex Action Recognition},
author    = {Hussein, Noureldien and Gavves, Efstratios and Smeulders, Arnold W.M.},
booktitle = {CVPR},
year      = {2019}
}

CVPR, 2017

Noureldien Hussein, Efstratios Gavves, Arnold W. M. Smeulders

A manifold is learned using contrastive loss, in which there is a joint embedding of videos of human events and their related articles. The result is end-to-end model with best results on TRECVID MED dataset.

@inproceedings{hussein2017unified,
title     = {Unified Embedding and Metric Learning for Zero-Exemplar Event Detection},
author    = {Hussein, Noureldien and Gavves, Efstratios and Smeulders, Arnold W.M.},
booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
year      = {2017}
}

TRECVID Workshop, 2016

Cees G. M. Snoek, Jianfeng Dong, Xirong Li, Xiaoxu Wang, Qijie Wei, Weiyu Lan, Efstratios Gavves, Noureldien Hussein, Dennis C. Koelma, Arnold W. M. Smeulders

Summary for our method used in the annual competition for zero-shot event recognition TRECVID MED 2016.

@inproceedings{snoek2016trecvid,
author    = {Snoek, Cees G. M. and Dong, Jianfeng and Li, Xirong and Wei, Qijie and Wang, Xiaoxu and Lan, Weiyu and Gavves, Efstratios and Hussein, Noureldien and Koelma, Dennis C. and Smeulders, Arnold W.M.},
title     = {University of Amsterdam and Renmin University at TRECVID 2016: Searching Video, Detecting Events and Describing Video},
booktitle = {TRECVID Workshop},
year      = {2016},
url       = {https://ivi.fnwi.uva.nl/isis/publications/2016/SnoekPTRECVID2016},
}