Jean-Philippe Thiran

Chat with Graph Search

Ask any question about EPFL courses, lectures, exercises, research, news, etc. or try the example questions below.

DISCLAIMER: The Graph Chatbot is not programmed to provide explicit or categorical answers to your questions. Rather, it transforms your questions into API requests that are distributed across the various IT services officially administered by EPFL. Its purpose is solely to collect and recommend relevant references to content that you can explore to help you answer your questions.

We propose to map the fast iterative shrinkage-thresholding algorithm to a deep neural network (DNN), with a sparsity prior in a concatenation of wavelet bases, in the context of compressive imaging. We exploit the DNN architecture to learn the optimal weight matrix of the corresponding reweighted l1-minimization problem. We later use the learned weight matrix for the image reconstruction process, which is recast as a simple l1-minimization problem. The approach, denoted as learned extended FISTA, shows pro

Detecting faces in images is a key step in numerous computer vision applications as face recognition for example. Face detection is a difficult task in image analysis because of the large face intra-class variability which is due to the important influence of the environmental conditions on the face aspect. The existing methods for face detection can be divided into holistic methods and feature based methods. We propose a new method for detecting frontal faces in complex images featuring two main contributi

Most recent test-time adaptation methods focus on only classification tasks, use specialized network architectures, destroy model calibration or rely on lightweight information from the source domain. To tackle these issues, this paper proposes a novel Tes ...

Despite significant progress toward super resolving more realistic images by deeper convolutional neural networks (CNNs), reconstructing fine and natural textures still remains a challenging problem. Recent works on single image super resolution (SISR) are ...

In the context of next-generation radio interferometers, we are facing a big challenge of how to economically process data. The classical dimensionality reduction technique, averaging visibilities on time, may dilute fast radio transients (FRT). We propose ...

A sound that we hear in a natural setting allows us to identify the sound source and localize it in space. The two aspects can be disrupted independently as shown in a study of 15 patients with focal right-hemispheric lesions. Four patients were normal in sound recognition but severely impaired in sound localization, whereas three other patients had difficulties in recognizing sounds but localized them well. The lesions involved the inferior parietal and frontal cortices, and the superior temporal gyrus in

MR connectomics is an emerging framework in neuro-science that combines diffusion MRI and whole brain tractography methodologies with the analytical tools of network science. In the present work we review the current methods enabling structural connectivity mapping with MRI and show how such data can be used to infer new information of both brain structure and function. We also list the technical challenges that should be addressed in the future to achieve high-resolution maps of structural connectivity. Fr

We advocate the use of a novel compressed sensing technique for accelerating the magnetic resonance image acquisition process, coined spread spectrum MR imaging or simply s2MRI. The method resides in pre-modulating the signal of interest by a linear chirp, ...

Ultrafast ultrasound has recently emerged as an alternative to traditional focused ultrasound. By virtue of the low number of insonifications it requires, ultrafast ultrasound enables the imaging of the human body at potentially very high frame rates. Howe ...

Intracranial aneurysms are a common pathologic condition with a potential severe complication: rupture. Effective treatment options exist, neurosurgical clipping and endovascular techniques, but guidelines for treatment are unclear and focus mainly on pati ...

Super-resolution (SR) has traditionally been based on pairs of high-resolution images (HR) and their low-resolution (LR) counterparts obtained artificially with bicubic downsampling. However, in real-world SR, there is a large variety of realistic image de ...

We present a method for dynamically integrating audio-visual information for speech recognition, based on the estimated reliability of the audio and visual streams. Our method uses an information theoretic measure, the entropy derived from the state probab ...

Diffusion-Weighted Magnetic Resonance Imaging is the only non-invasive technique available to infer the underlying brain tissue microstructure. Currently, one of the promising methods for microstructure imaging is signal modelling using convex formulation, e.g. using the COMMIT framework. Despite the benefits introduced with such a framework, an important limitation is the long convergence time, making the method unappealing for clinical applications. In order to address this limitation, we propose to use a

Training robust deep learning (DL) systems for medical image classification or segmentation is challenging due to limited images covering different disease types and severity. We propose an active learning (AL) framework to select most informative samples ...

There is growing interest in understanding the role of the non-injured contra-lateral hemisphere in stroke recovery. In the experimental field, histological evidence has been reported that structural changes occur in the contra-lateral connectivity and cir ...

Cardiac arrhythmias present a significant global health concern. The advent of wearable devices utilizing photoplethysmography gives the opportunity to screen large populations, hence offering the potential for early detection of pathological rhythms and r ...

The exploration of the human connectome, a term denoting the global structural connectivity of the brain, is accessible to MRI at millimeter and centimeter scales. In this paper, we propose a methodology to map the connectome by constructing normalized who ...

In unfocused ultrasound imaging, a delay-and-sum algorithm is commonly used to reconstruct one image per emission. When multiple emissions are performed, individual images can be combined by coherent compounding to improve image quality. Alternative method ...

In this paper we deal with the multi-object tracking problem in the particular case of pedestrians, assuming the detection step already done. We use a Bayesian framework to combine the likelihood term provided by an image correlation algorithm with a prior distribution given by a discrete choice model for pedestrian behavior, calibrated on real data. We aim to show how the combination of the image information with a model of pedestrian behavior can provide appreciable results in real and complex scenarios.

Purpose Biophysical models of diffusion MRI have been developed to characterize microstructure in various tissues, but existing models are not suitable for tissue composed of permeable spherical cells. In this study we introduce Cellular Exchange Imaging ( ...

Jean-Philippe Thiran

Chat with Graph Search

Learning the weight matrix for sparsity averaging in compressive imaging

Face Detection with Mixtures of Boosted Discriminant Features

TeSLA: Test-Time Self-Learning With Automatic Adversarial Augmentation

Benefiting from Multitask Learning to Improve Single Image Super-Resolution

Fourier dimensionality reduction for fast radio transients

What and where in human audition: Distinct cortical processing pathways revealed by fMRI

MR connectomics: Principles and challenges

Accelerated MR imaging with spread spectrum encoding

Windowed Radon Transform and Tensor Rank-1 Decomposition for Adaptive Beamforming in Ultrafast Ultrasound

Intracranial Aneurysms: Wall Motion Analysis for Prediction of Rupture

Benefiting From Bicubically Down-Sampled Images for Learning Real-World Image Super-Resolution

Using entropy as a stream reliability estimate for audio-visual speech recognition

Learning global brain microstructure maps using trainable sparse encoders

Efficient Active Learning for Image Classification and Segmentation using a Sample Selection and Conditional Generative Adversarial Network

Diffusion Spectrum Imaging after stroke shows structural changes in the contra-lateral motor network correlating with functional recovery

A Comparative Study on Detecting Heart Beats in Photoplethysmography Signals in Presence of Various Cardiac Arrhythmias

Diffusion MR Image Segmentation: Towards Global Brain Connectivity Analysis

Inverse Problems With Multiple Plane Waves: The Angular Simplification

Bayesian Integration of a Discrete Choice Pedestrian Behavioral Model and Image Correlation Techniques for Automatic Multi Object Tracking

Cellular EXchange Imaging (CEXI): Evaluation of a diffusion model including water exchange in cells using numerical phantoms of permeable spheres