Publications de Pascal Fua | EPFL Graph Search

Chattez avec Graph Search

Posez n’importe quelle question sur les cours, conférences, exercices, recherches, actualités, etc. de l’EPFL ou essayez les exemples de questions ci-dessous.

AVERTISSEMENT : Le chatbot Graph n'est pas programmé pour fournir des réponses explicites ou catégoriques à vos questions. Il transforme plutôt vos questions en demandes API qui sont distribuées aux différents services informatiques officiellement administrés par l'EPFL. Son but est uniquement de collecter et de recommander des références pertinentes à des contenus que vous pouvez explorer pour vous aider à répondre à vos questions.

The goal of Deep Domain Adaptation is to make it possible to use Deep Nets trained in one domain where there is enough annotated training data in another where there is little or none. Most current approaches have focused on learning feature representation ...

We have developed a stereo algorithm that relies on grey level correlation followed by interpolation using an energy based technique. During the correlation phase the two images play a symmetric role and we use a validity criterion for the matches that eliminates the gross errors: when the images cannot be correlated reliably, due to lack of texture or occlusions for example, the algorithm does not produce wrong matches but a very sparse disparity map as opposed to a dense one when the correlation is succes

Reconstructing 3D shape from 2D sketches has long been an open problem because the sketches only provide very sparse and ambiguous information. In this paper, we use an encoder/decoder architecture for the sketch to mesh translation. When integrated into a ...

Diffusion models generating images conditionally on text, such as Dall-E 2 [51] and Stable Diffusion[53], have recently made a splash far beyond the computer vision community. Here, we tackle the related problem of generating point clouds, both uncondition ...

We present a new approach to matching graphs embedded in R2 or R3. Unlike earlier methods, our approach does not rely on the similarity of local appearance features, does not require an initial alignment, can handle partial matches, and can cope with non-l ...

The present invention generally relates to a method, a system and a computer program for shape optimisation of a technical device adapted to be exposed to a fluid flowing around a contour of said device with respect to its fluid dynamic parameters. In orde ...

We present a new approach for the automated segmentation of synapses in image stacks acquired by Electron Microscopy (EM) that relies on image features specifically designed to take spatial context into account. These features are used to train a classifie ...

By contrast, in this paper, we use temporal motion models based on Principal Component Analysis to formulate the tracking problem as one of minimizing differentiable objective functions. The differential structure of these functions is rich enough to yield good convergence properties using a deterministic optimization scheme at a much reduced computational cost. Furthermore, by using a multi-activity database, we can partially overcome one of the major limitations of approaches that rely on motion models, n

Electron and Light Microscopy imaging can now deliver high-quality image stacks of neural structures. However, the amount of human annotation effort required to analyze them remains a major bottleneck. While Machine Learning algorithms can be used to help ...

Multi-object tracking can be achieved by detecting objects in individ- ual frames and then linking detections across frames. Such an approach can be made very robust to the occasional detection failure: If an object is not detected in a frame but is in previous and following ones, a cor- rect trajectory will nevertheless be produced. By contrast, a false-positive detection in a few frames will be ignored. However, when dealing with a multiple target problem, the linking step results in a difficult optimizat

We propose a fast approach to 3D object detection and pose estimation that owes its robustness to a training phase during which the target object slowly moves with respect to the camera. No additional information is provided to the system, save a very roug ...

Multi-object tracking can be achieved by detecting objects in individual frames and then linking detections across frames. Such an approach can be made very robust to the occasional detection failure: If an object is not detected in a frame but is in previous and following ones, a correct trajectory will nevertheless be produced. By contrast, a false-positive detection in a few frames will be ignored. However, when dealing with a multiple target problem, the linking step results in a difficult optimization

This protocol describes how dendrites and axons, imaged in vivo, can subsequently be analyzed in 3D using focused ion beam scanning electron microscopy (FIBSEM). The fluorescent structures are identified after chemical fixation and their position highlight ...

Using silhouettes in uncontrolled environments typically requires handling occlusions as well as changing or cluttered backgrounds, which limits the applicability of most silhouette based methods. For the purpose of 3--D shape modeling, we show that repres ...

A method for enforcing smoothness constraints on surface meshes produced by a Graph Convolutional Neural Network (GCNN) including the steps of reading image data from a memory, the image data including two-dimensional image data representing a three-dimens ...

CNN-based volumetric methods that label individual voxels now dominate the field of biomedical segmentation. However, 3D surface representations are often required for proper analysis. They can be obtained by post-processing the labeled volumes which typic ...

Deep networks such as the U-Net are outstanding at segmenting biomedical images when enough training data is available, but only then. Here we introduce a Domain Adaptation approach that relies on two coupled U-Nets that either regularize or share correspo ...

Many biological and medical tasks require the delineation of 3D curvilinear structures such as blood vessels and neurites from image volumes. This is typically done using neural networks trained by minimizing voxel-wise loss functions that do not capture t ...

Aerodynamic shape optimization (ASO) plays a crucial role in aerodynamic design by enhancing an object’s aerodynamic performance while satisfying design constraints. Traditional ASO is costly and time-intensive, requiring extensive manual tuning to configure parameterization methods that define the design space and manipulate the shape geometry. These conventional parameterization methods limit flexibility and require extensive domain prior knowledge on the design tasks and geometric processing, making effi

Pascal Fua

Chattez avec Graph Search

Residual Parameter Transfer for Deep Domain Adaptation

A Parallel Stereo Algorithm that Produces Dense Depth Maps and Preserves Image Features

Sketch2Mesh: Reconstructing and Editing 3D Shapes from Sketches

GECCO: Geometrically-Conditioned Point Diffusion Models

Robust Non-Rigid Registration of 2D and 3D Graphs

Shape optimisation of technical devices via gradient descent using convolutional neural network proxies

Learning Context Cues for Synapse Segmentation

3D Human Body Tracking using Deterministic Temporal Motion Models

Domain Adaptation for Microscopy Imaging

Multiple Object Tracking using Flow Linear Programming

Feature Harvesting for Tracking-by-Detection

Multiple Object Tracking using Flow Linear Programming

Correlative In Vivo 2-Photon Beam Scanning Electron Microscopy: 3D Analysis of Neuronal Ultrastructure

Implicit Meshes for Effective Silhouette Handling

A method and system for enforcing smoothness constraints on surface meshes from a graph convolutional neural network

Voxel2Mesh: 3D Mesh Model Generation from Volumetric Data

Multiple Object Tracking using K-Shortest Paths Optimization

A Domain-Adaptive Two-Stream U-Net For Electron Microscopy Image Segmentation

Enforcing Connectivity of 3D Linear Structures Using Their 2D Projections

DeepGeo: Deep Geometric Mapping for Automated and Effective Parameterization in Aerodynamic Shape Optimization