Search Results for author: Jan van Gemert

Found 28 papers, 10 papers with code

Heuristics2Annotate: Efficient Annotation of Large-Scale Marathon Dataset For Bounding Box Regression

no code implementations6 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.

Person Re-Identification

Spectral Leakage and Rethinking the Kernel Size in CNNs

1 code implementation25 Jan 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.

Deep Continuous Networks

no code implementations1 Jan 2021 Nergis Tomen, Silvia Laura Pintea, Jan van Gemert

Finally, we suggest that the continuous representations learned by DCNs may enable computationally efficient implementations.

Less bits is more: How pruning deep binary networks increases weight capacity

no code implementations1 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.

Real-time Webcam Heart-Rate and Variability Estimation with Clean Ground Truth for Evaluation

no code implementations31 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.

Heart Rate Variability

Deep Unsupervised Image Hashing by Maximizing Bit Entropy

1 code implementation22 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.

Semantic Retrieval

Zoom-CAM: Generating Fine-grained Pixel Annotations from Image Labels

no code implementations16 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.

Weakly-Supervised Object Localization Weakly-Supervised Semantic Segmentation

Respecting Domain Relations: Hypothesis Invariance for Domain Generalization

no code implementations15 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.

Domain Generalization

WeightAlign: Normalizing Activations by Weight Alignment

no code implementations14 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.

Domain Adaptation Semantic Segmentation

Efficiency in Real-time Webcam Gaze Tracking

no code implementations2 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.

Black Magic in Deep Learning: How Human Skill Impacts Network Training

1 code implementation13 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.

Hyperparameter Optimization

Evaluating the performance of the LIME and Grad-CAM explanation methods on a LEGO multi-label image classification task

no code implementations4 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.

Multi-Label Image Classification

Rethinking Online Action Detection in Untrimmed Videos: A Novel Online Evaluation Protocol

1 code implementation26 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.

Action Detection

Attention-Aware Age-Agnostic Visual Place Recognition

no code implementations11 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.

Domain Adaptation Image Retrieval +1

Running Event Visualization using Videos from Multiple Cameras

no code implementations6 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.

Person Re-Identification Scene Text +2

Push for Quantization: Deep Fisher Hashing

no code implementations31 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.

Quantization Semantic Similarity +1

Deep Visual City Recognition Visualization

1 code implementation6 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.

Image Classification Semantic Segmentation

ViDeNN: Deep Blind Video Denoising

1 code implementation24 Apr 2019 Michele Claus, Jan van Gemert

We propose ViDeNN: a CNN for Video Denoising without prior knowledge on the noise distribution (blind denoising).

Denoising Video Denoising

Object-Extent Pooling for Weakly Supervised Single-Shot Localization

no code implementations19 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.

Region Proposal Weakly-Supervised Object Localization

One-Step Time-Dependent Future Video Frame Prediction with a Convolutional Encoder-Decoder Neural Network

no code implementations14 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.

Optical Flow Estimation

Tubelets: Unsupervised action proposals from spatiotemporal super-voxels

no code implementations7 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.

Action Localization

Structured Receptive Fields in CNNs

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.

Action Localization with Tubelets from Motion

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.

Action Localization

Cannot find the paper you are looking for? You can Submit a new open access paper.