no code implementations • 16 Apr 2024 • Zhi-Yi Lin, Jouh Yeong Chew, Jan van Gemert, Xucong Zhang
We propose an end-to-end approach for gaze target detection: predicting a head-target connection between individuals and the target image regions they are looking at.
no code implementations • 31 Jan 2024 • Sven de Witte, Ombretta Strafforello, Jan van Gemert
Bounding boxes are often used to communicate automatic object detection results to humans, aiding humans in a multitude of tasks.
no code implementations • 3 Nov 2023 • Xinqi Li, Yi Zhang, Yidong Zhao, Jan van Gemert, Qian Tao
To address the challenge, we propose a novel motion correction framework based on robust principle component analysis (rPCA) that decomposes quantitative cardiac MRI into low-rank and sparse components, and we integrate the groupwise CNN-based registration backbone within the rPCA framework.
1 code implementation • 6 Oct 2023 • Athanasios Masouris, Jan van Gemert
Chess recognition is the task of extracting the chess piece configuration from a chessboard image.
1 code implementation • 12 Sep 2023 • Alessandro Duico, Ombretta Strafforello, Jan van Gemert
To this end, we curate a large video benchmark, the YTMR500 dataset, which comprises 500 YouTube videos with MR data annotations.
1 code implementation • 8 Sep 2023 • Casper van Engelenburg, Seyran Khademi, Jan van Gemert
In this paper, an effective evaluation metric for judging the structural similarity of floor plans, coined SSIG (Structural Similarity by IoU and GED), is proposed based on both image and graph distances.
no code implementations • 24 Aug 2023 • Jan Warchocki, Teodor Oprescu, Yunhan Wang, Alexandru Damacus, Paul Misterka, Robert-Jan Bruintjes, Attila Lengyel, Ombretta Strafforello, Jan van Gemert
This work explores and measures how current deep temporal action localization models perform in settings constrained by the amount of data or computational power.
1 code implementation • 22 Aug 2023 • Ombretta Strafforello, Klamer Schutte, Jan van Gemert
In the current deep learning paradigm for automatic action recognition, it is imperative that models are trained and tested on datasets and tasks that evaluate if such models actually learn and reason over long-term information.
1 code implementation • 22 Aug 2023 • Ombretta Strafforello, Xin Liu, Klamer Schutte, Jan van Gemert
Previous work on long-term video action recognition relies on deep 3D-convolutional models that have a large temporal receptive field (RF).
no code implementations • 22 Aug 2023 • Tom Edixhoven, Attila Lengyel, Jan van Gemert
In this paper we show how Group Equivariant Convolutional Neural Networks use subsampling to learn to break equivariance to their symmetries.
no code implementations • 31 May 2023 • Robert-Jan Bruintjes, Attila Lengyel, Marcos Baptista Rios, Osman Semih Kayhan, Davide Zambrano, Nergis Tomen, Jan van Gemert
The third edition of the "VIPriors: Visual Inductive Priors for Data-Efficient Deep Learning" workshop featured four data-impaired challenges, focusing on addressing the limitations of data availability in training deep learning models for computer vision tasks.
no code implementations • 5 Apr 2023 • Robert-Jan Bruintjes, Tomasz Motyka, Jan van Gemert
We therefore investigate what can increase the learned equivariance in neural networks, and find that data augmentation, reduced model capacity and inductive bias in the form of convolutions induce higher learned equivariance in neural networks.
1 code implementation • 4 Mar 2023 • Joris Quist, Yunqiang Li, Jan van Gemert
Our analysis makes it possible to understand how magnitude-based hyperparameters influence the training of binary networks which allows for new optimization filters specifically designed for binary neural networks that are independent of their real-valued interpretation.
1 code implementation • 13 Jan 2023 • Marian Bittner, Wei-Tse Yang, Xucong Zhang, Ajay Seth, Jan van Gemert, Frans C. T. van der Helm
Markerless estimation of 3D Kinematics has the great potential to clinically diagnose and monitor movement disorders without referrals to expensive motion capture labs; however, current approaches are limited by performing multiple de-coupled steps to estimate the kinematics of a person from videos.
1 code implementation • 25 Nov 2022 • Liang Zeng, Attila Lengyel, Nergis Tömen, Jan van Gemert
For unsupervised semantic segmentation of urban scenes, our method surpasses the previous state-of-the-art baseline by +7. 14% in mIoU on Cityscapes and +6. 65% on KITTI.
1 code implementation • 25 Oct 2022 • Sieger Falkena, Hadi Jamali-Rad, Jan van Gemert
Binary Neural Networks (BNNs) are receiving an upsurge of attention for bringing power-hungry deep learning towards edge devices.
1 code implementation • 29 Aug 2022 • Ruben Sangers, Jan van Gemert, Sander van Cranenburgh
However, the blackbox nature of deep learning models hampers urban planners to understand what landscape objects contribute to a particularly high quality or low quality urban space perception.
1 code implementation • 4 Aug 2022 • Yeshwanth Napolean, Anwesh Marwade, Nergis Tomen, Puck Alkemade, Thijs Eijsvogels, Jan van Gemert
Existing work can robustly measure heart rate under some degree of motion by face tracking.
1 code implementation • 28 Jul 2022 • Ombretta Strafforello, Vanathi Rajasekart, Osman S. Kayhan, Oana Inel, Jan van Gemert
Our work is the first to evaluate IoU with humans and makes it clear that relying on IoU scores alone to evaluate localization errors might not be sufficient.
no code implementations • 1 Jun 2022 • Amogh Gudi, Fritjof Büttner, Jan van Gemert
Mean squared error (MSE) is one of the most widely used metrics to expression differences between multi-dimensional entities, including images.
1 code implementation • 30 Mar 2022 • Burak Yildiz, Seyran Khademi, Ronald Maria Siebes, Jan van Gemert
Our result credits the best accuracy to the ResNet-101 model pre-trained on the Landmarks dataset for both verification and retrieval tasks by 84% and 24%, respectively.
Ranked #1 on Image Classification on AmsterTime (using extra training data)
no code implementations • 21 Jan 2022 • Attila Lengyel, Robert-Jan Bruintjes, Marcos Baptista Rios, Osman Semih Kayhan, Davide Zambrano, Nergis Tomen, Jan van Gemert
The second edition of the "VIPriors: Visual Inductive Priors for Data-Efficient Deep Learning" challenges featured five data-impaired challenges, where models are trained from scratch on a reduced number of training samples for various key computer vision tasks.
no code implementations • 23 Dec 2021 • Yancong Lin, Silvia-Laura Pintea, Jan van Gemert
Experiments on both synthetic and real-world datasets show the benefit of our proposed changes for improved data efficiency and inference speed.
no code implementations • 23 Oct 2021 • Tuhin Das, Robert-Jan Bruintjes, Attila Lengyel, Jan van Gemert, Sara Beery
While domain adaptation is generally applied on completely synthetic source domains and real target domains, we explore how domain adaptation can be applied when only a single rare class is augmented with simulated samples.
no code implementations • 9 Jun 2021 • Yancong Lin, Silvia-Laura Pintea, Jan van Gemert
Current work on lane detection relies on large manually annotated datasets.
no code implementations • 6 Apr 2021 • Pranjal Singh Rajput, Yeshwanth Napolean, Jan van Gemert
Additionally, due to crowdedness and occlusion in the videos, aligning the identity of runners across multiple disjoint cameras is a challenge.
1 code implementation • 5 Mar 2021 • Robert-Jan Bruintjes, Attila Lengyel, Marcos Baptista Rios, Osman Semih Kayhan, Jan van Gemert
We present the first edition of "VIPriors: Visual Inductive Priors for Data-Efficient Deep Learning" challenges.
1 code implementation • ICCV 2021 • Nergis Tomen, Jan van Gemert
We show that the small size of CNN kernels make them susceptible to spectral leakage, which may induce performance-degrading artifacts.
no code implementations • 1 Jan 2021 • Yunqiang Li, Silvia Laura Pintea, Jan van Gemert
We make the observation that pruning weights adds the value 0 as an additional symbol and thus increases the information capacity of the network.
no code implementations • 31 Dec 2020 • Amogh Gudi, Marian Bittner, Jan van Gemert
We introduce a refined and efficient real-time rPPG pipeline with novel filtering and motion suppression that not only estimates heart rates, but also extracts the pulse waveform to time heart beats and measure heart rate variability.
1 code implementation • 22 Dec 2020 • Yunqiang Li, Jan van Gemert
This layer is shown to minimize a penalized term of the Wasserstein distance between the learned continuous image features and the optimal half-half bit distribution.
1 code implementation • 1 Dec 2020 • Burak Yildiz, Hayley Hung, Jesse H. Krijthe, Cynthia C. S. Liem, Marco Loog, Gosia Migut, Frans Oliehoek, Annibale Panichella, Przemyslaw Pawelczak, Stjepan Picek, Mathijs de Weerdt, Jan van Gemert
We present ReproducedPapers. org: an open online repository for teaching and structuring machine learning reproducibility.
no code implementations • 16 Oct 2020 • Xiangwei Shi, Seyran Khademi, Yunqiang Li, Jan van Gemert
Current weakly supervised object localization and segmentation rely on class-discriminative visualization techniques to generate pseudo-labels for pixel-level training.
no code implementations • 15 Oct 2020 • Ziqi Wang, Marco Loog, Jan van Gemert
In this work, we define DIRs employed by existing works in probabilistic terms and show that by learning DIRs, overly strict requirements are imposed concerning the invariance.
no code implementations • 14 Oct 2020 • Xiangwei Shi, Yunqiang Li, Xin Liu, Jan van Gemert
Such methods are less stable than BN as they critically depend on the statistics of a single input sample.
no code implementations • 2 Sep 2020 • Amogh Gudi, Xin Li, Jan van Gemert
To do so, we evaluate the computational speed/accuracy trade-off for the CNN and the calibration effort/accuracy trade-off for screen calibration.
1 code implementation • 13 Aug 2020 • Kanav Anand, Ziqi Wang, Marco Loog, Jan van Gemert
Our study investigates the subjective human factor in comparisons of state of the art results and scientific reproducibility in deep learning.
no code implementations • 4 Aug 2020 • David Cian, Jan van Gemert, Attila Lengyel
In this paper, we run two methods of explanation, namely LIME and Grad-CAM, on a convolutional neural network trained to label images with the LEGO bricks that are visible in them.
1 code implementation • 26 Mar 2020 • Marcos Baptista Rios, Roberto J. López-Sastre, Fabian Caba Heilbron, Jan van Gemert, F. Javier Acevedo-Rodríguez, S. Maldonado-Bascón
Our results confirm the problems of the previous evaluation protocols, and suggest that an IA-based protocol is more adequate to the online scenario.
1 code implementation • 22 Mar 2020 • Marcos Baptista Rios, Roberto J. López-Sastre, Fabian Caba Heilbron, Jan van Gemert, Francisco Javier Acevedo-Rodríguez, Saturnino Maldonado-Bascón
The problem of Online Human Behaviour Recognition in untrimmed videos, aka Online Action Detection (OAD), needs to be revisited.
no code implementations • 11 Sep 2019 • Ziqi Wang, Jiahui Li, Seyran Khademi, Jan van Gemert
Different from conventional VPR settings where the query images and gallery images come from the same domain, we propose a more common but challenging setup where the query images are collected under a new unseen condition.
no code implementations • 6 Sep 2019 • Yeshwanth Napolean, Priadi Teguh Wibowo, Jan van Gemert
To this end, we identify two methods for runner identification at different points of the event, for determining their trajectory.
no code implementations • 3 Sep 2019 • Amogh Gudi, Marian Bittner, Roelof Lochmans, Jan van Gemert
Remote photo-plethysmography (rPPG) uses a remotely placed camera to estimating a person's heart rate (HR).
no code implementations • 31 Aug 2019 • Yunqiang Li, Wenjie Pei, Yufei zha, Jan van Gemert
In this paper we push for quantization: We optimize maximum class separability in the binary space.
1 code implementation • 6 May 2019 • Xiangwei Shi, Seyran Khademi, Jan van Gemert
(ii) A pretrained semantic segmentation model is used to label objects in pixel level, and then we introduce statistical measures to quantitatively evaluate the interpretability of discriminate objects.
1 code implementation • 24 Apr 2019 • Michele Claus, Jan van Gemert
We propose ViDeNN: a CNN for Video Denoising without prior knowledge on the noise distribution (blind denoising).
Ranked #2 on Color Image Denoising on CBSD68 sigma10
no code implementations • 19 Jul 2017 • Amogh Gudi, Nicolai van Rosmalen, Marco Loog, Jan van Gemert
To facilitate this, we propose a novel global pooling technique called Spatial Pyramid Averaged Max (SPAM) pooling for training this CAM-based network for object extent localisation with only weak image-level supervision.
no code implementations • 14 Feb 2017 • Vedran Vukotić, Silvia-Laura Pintea, Christian Raymond, Guillaume Gravier, Jan van Gemert
There is an inherent need for autonomous cars, drones, and other robots to have a notion of how their environment behaves and to anticipate changes in the near future.
no code implementations • 7 Jul 2016 • Mihir Jain, Jan van Gemert, Hervé Jégou, Patrick Bouthemy, Cees G. M. Snoek
First, inspired by selective search for object proposals, we introduce an approach to generate action proposals from spatiotemporal super-voxels in an unsupervised manner, we call them Tubelets.
3 code implementations • CVPR 2016 • Jörn-Henrik Jacobsen, Jan van Gemert, Zhongyu Lou, Arnold W. M. Smeulders
We combine these ideas into structured receptive field networks, a model which has a fixed filter basis and yet retains the flexibility of CNNs.
no code implementations • CVPR 2014 • Mihir Jain, Jan van Gemert, Herve Jegou, Patrick Bouthemy, Cees G. M. Snoek
Our approach significantly outperforms the state-of-the-art on both datasets, while restricting the search of actions to a fraction of possible bounding box sequences.