Publications by Pascal Fua | EPFL Graph Search

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.

This work collects the explorations conducted within the EPFL+ECAL Lab by several designers to interpret the various spheres of action of Augmented Reality in order to derive visual principles. These principles seek to contribute to developing a specific v ...

State-of-the-art semantic or instance segmentation deep neural networks (DNNs) are usually trained on a closed set of semantic classes. As such, they are ill-equipped to handle previously-unseen objects. However, detecting and localizing such objects is cr ...

Immense amounts of high resolution data are now routinely produced thanks to recent advances in EM imaging. While a strong demand for automated analysis now exists, it is stifled by the lack of robust automatic 3D segmentation techniques. State-of-the-art Computer Vision algorithms designed to operate on natural 2D images tend to perform poorly when applied to EM image stacks for a number of reasons. The sheer size of a typical EM image stack renders many segmentation schemes intractable. Most approaches re

3--D detection and shape recovery of a non-rigid surface from video sequences require deformation models to effectively take advantage of potentially noisy image data. Here we introduce an approach to creating such models for deformable 3--D surfaces. We e ...

Learning general image representations has proven key to the success of many computer vision tasks. For example, many approaches to image understanding problems rely on deep networks that were initially trained on ImageNet, mostly because the learned featu ...

e propose a deformable surface parameterization that is generic and lets us automatically build registered shape databases. This allows us to directly derive low-dimensional shape models using a simple dimensionality reduction technique. This addresses one of the biggest difficulties in example-based shape modeling: Building the required database, which is often a difficult and painstaking process. We incorporate the resulting models into a monocular tracking system that we use to capture the complex deform

We introduce a novel behavioral model to describe pedestrians motions, which is able to capture sophisticated motion patterns resulting from the mixture of different categories of random trajectories. Due to its simplicity, this model can be learned from video sequences in a totally unsupervised manner through an Expectation-Maximization procedure. When integrated into a complete multi-camera tracking system, it improves the tracking performance in ambiguous situations, compared to a standard ad-hoc isotrop

We propose an approach to gait analysis that relies on fitting 3--D temporal motion models to synchronized video sequences. These models allow us not only to track but also to recover motion parameters that can be used to recognize people and characterize their style. Because our method is robust to occlusions and insensitive to changes in direction of motion, our proposed approach has the potential to overcome some of the main limitations of current gait analysis methods. This is an important step towards

We propose a robust and accurate method to extract the centerlines and scale of tubular structures in 2D images and 3D volumes. Existing techniques rely either on filters designed to respond to ideal cylindrical structures, which lose accuracy when the lin ...

In this paper, we evaluate the Probabilistic Occupancy Map (POM) pedestrian detection algorithm on the PETS 2009 benchmark dataset. POM is a multi-camera generative detection method, which estimates ground plane occupancy from multiple background subtraction views. Occupancy probabilities are iteratively estimated by fitting a synthetic model of the background subtraction to the binary foreground motion. Furthermore, we test the integration of this algorithm into a larger framework designed for understandin

Generating detailed simpliﬁed building models such as the ones present on Google Earth is often a difﬁcult and lengthy manual task, requiring advanced CAD software and a combination of ground imagery, LIDAR data and blueprints. Nowadays, UAVs such as the Falcon 8 have reached the maturity to offer an affordable, fast and easy way to capture large amounts of oblique images covering all part of a building. In this paper we present a state-of-the-art photogrammetry and visual reconstruction pipeline provided b

Machine Learning techniques play an increasingly vital role in the analysis of Biomedical imagery, as in all other areas of Computer Vision. However, in this specific context, they suffer from the fact that experimental conditions and protocols change often and that acquiring sufficient amounts of new training data after each image acquisition is impractical. In this paper, we propose an effective method to train a non-linear SVM using a very small amount of new data by leveraging data obtained under differ

LiftPose3D infers three-dimensional poses from two-dimensional data or from limited three-dimensional data. The approach is illustrated for videos of behaving Drosophila, mice, rats and macaques. ...

Our goal is to capture the pose of neuroscience model organisms, without using any manual supervision, to be able to study how neural circuits orchestrate behaviour. Human pose estimation attains remarkable accuracy when trained on real or simulated datase ...

Unconstrained Least-Squares minimization is a well-studied problem. For example, the Levenberg-Marquardt is extremely effective and numerous implementations are readily available. These algorithms are, however, not designed to perform least-squares minimization under hard constraints. This short report outlines two very simple approaches to doing this. The first relies on standard Lagrange multipliers. The second is inspired by inverse kinematics techniques.

Analyzing and training 3D body posture models depend heavily on the availability of joint labels that are commonly acquired through laborious manual annotation of body joints or via marker-based joint localization using carefully curated markers and capturing systems. However, such annotations are not always available, especially for people performing unusual activities. In this paper, we propose an algorithm that learns to discover 3D keypoints on human bodies from multiple-view images without any supervis

We present a novel approach to 3D delineation of dendritic networks in noisy image stacks. We achieve a level of automation beyond that of state-of-the-art systems, which model dendrites as continuous tubular structures and postulate simple appearance mode ...

Recovering the 3D shape of deformable surfaces from single images is difficult because many different shapes have very similar projections. This is commonly addressed by restricting the set of possible shapes to linear combinations of deformation modes and ...

We propose a novel, connectivity-oriented loss function for training deep convolutional networks to reconstruct network-like structures, like roads and irrigation canals, from aerial images. The main idea behind our loss is to express the connectivity of r ...

State-of-the-art approaches to detecting ridge-like structures in images rely on filters designed to respond to locally linear intensity features. While these approaches may be optimal for ridges whose appearance is close to being ideal, their performance degrades quickly in the presence of structured noise that corrupts the image signal, potentially to the point where it truly does not conform to the ideal model anymore. In this paper, we address this issue by introducing a learning framework that relies o

Pascal Fua

Chat with Graph Search

Turning Augmented Reality into a media: Design exploration to build a dedicated visual language

SegmentMeIfYouCan: A Benchmark for Anomaly Segmentation

Supervoxel-Based Segmentation of EM Image Stacks with Learned Shape Features

Surface Deformation Models for Non-Rigid 3--D Shape Recovery

Neural Scene Decomposition for Multi-Person Motion Capture

Parameterizing Deformable Surfaces for Monocular 3--D Tracking

Multi-Camera Tracking and Atypical Motion Detection with Behavioral Maps

3D Tracking for Gait Characterization and Recognition

Multiscale Centerline Detection by Learning a Scale-Space Distance Transform

Evaluation of Probabilistic Occupancy Map People Detection for Surveillance Systems

Simplified Building Models Extraction From Ultra-Light UAV Imagery

Transfer Learning by Sharing Support Vectors

LiftPose3D, a deep learning-based approach for transforming two-dimensional to three-dimensional poses in laboratory animals

Deformation-aware Unpaired Image Translation for Pose Estimation on Laboratory Animals

Least-Squares Minimization Under Constraints

Unsupervised 3D Keypoint Discovery with Multi-View Geometry

Automated Delineation of Dendritic Networks in Noisy Image Stacks

Exploring Ambiguities for Monocular Non-Rigid Shape Estimation

Promoting Connectivity of Network-Like Structures by Enforcing Region Separation

Rotational Features Extraction for Ridge Detection