Publications by Mathieu Salzmann

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.

Accurately estimating model performance poses a significant challenge, particularly in scenarios where the source and target domains follow different data distributions. Most existing performance prediction methods heavily rely on the source data in their ...

Modern 3D human pose estimation techniques rely on deep networks, which require large amounts of training data. While weakly-supervised methods require less supervision, by utilizing 2D poses or multi-view imagery without annotations, they still need a suf ...

HOISDF: Constraining 3D Hand-Object Pose Estimation with Global Signed Distance Fields, CVPR 2024 Haozhe Qi, Chen Zhao, Mathieu Salzmann, Alexander Mathis. Affiliation: EPFL Date: June, 2024 Link to the CVPR article: https://openaccess.thecvf.com/conte ...

AI weather models and weather-based foundation models have demonstrated impressive skills in short- to medium-range forecasts. While most weather models become unstable on longer time scales, a wide variety of AI climate emulators have been proposed, raising questions about the fundamental differences between these approaches.In this work, we compare state-of-the-art models when producing rollouts on annual time scales. We quantify and characterize different types of instability: smoothing, visual artifacts

Recent years have seen the development of mature solutions for reconstructing deformable surfaces from a single image, provided that they are relatively well-textured. By contrast, recovering the 3D shape of texture-less surfaces remains an open problem, a ...

We present an Unsupervised Domain Adaptation strategy to compensate for domain shifts on Electron Microscopy volumes. Our method aggregates visual correspondences—motifs that are visually similar across different acquisitions—to infer changes on the parame ...

In this paper, we tackle the problem of static 3D cloth draping on virtual human bodies. We introduce a two-stream deep network model that produces a visually plausible draping of a template cloth on virtual 3D bodies by extracting features from both the b ...

Foundation models for the Earth system have gained popularity, as they are starting to surpass numerical solvers in the accuracy of predicting Earth’s condition while requiring fewer computational resources. The Earth System Foundation Model (ESFM) contributes to this research direction by further extending the foundation models' flexibility.The forecasting capabilities of ESFM are achieved in an autoregressive manner, using data from the t0 - Δt and t0 timesteps to produce a prediction for t0 + Δt. This ap

This paper assesses the feasibility of estimating the position and attitude of a target satellite using computer vision, not during proximity operations, but during conjunctions at distances greater than 1 km. The selected target is a small 5 × 5 × 5 cm3 P ...

In this work, we tackle the task of estimating the 6D pose of an object from point cloud data. While recent learning-based approaches have shown remarkable success on synthetic datasets, we have observed them to fail in the presence of real-world data. We ...

Weight sharing has become a de facto standard in neural architecture search because it enables the search to be done on commodity hardware. However, recent works have empirically shown a ranking disorder between the performance of stand-alone architectures ...

We analyze the influence of adversarial training on the loss landscape of machine learning models. To this end, we first provide analytical studies of the properties of adversarial loss functions under different adversarial budgets. We then demonstrate tha ...

Existing techniques to encode spatial invariance within deep convolutional neural networks (CNNs) apply the same warping field to all the feature channels. This does not account for the fact that the individual feature channels can represent different sema ...

We propose a new formulation to the non-rigid structure-from-motion problem that only requires the deforming surface to meaning that its differential structure is preserved. This is a much weaker assumption than the traditional ones of isometry or conforma ...

We address the problem of stably and efficiently training a deep neural network robust to adversarial perturbations bounded by an

l_1

norm. We demonstrate that achieving robustness against

l_1

-bounded perturbations is more challenging than in the

l_2

...

Most modern image-based 6D object pose estimation methods learn to predict 2D-3D correspondences, from which the pose can be obtained using a PnP solver. Because of the non-differentiable nature of common PnP solvers, these methods are supervised via the i ...

The Open Books is an interactive installation developed by the EPFL+ECAL Lab in collaboration with the Photo Elysée Museum designed to engage the public through interaction design to support the comprehension of photobooks. A photobook is a work in which photographic images are edited and arranged into a sequence, with meaning produced through editorial and layout relations rather than through isolated images or text. While photobooks are increasingly mediated through exhibitions and interactive museum inst

Accurate 3D human pose estimation from single images is possible with sophisticated deep-net architectures that have been trained on very large datasets. However, this still leaves open the problem of capturing motions for which no such database exists. Ma ...

Human motion prediction, the task of predicting future 3D human poses given a sequence of observed ones, has been mostly treated as a deterministic problem. However, human motion is a stochastic process: Given an observed sequence of poses, multiple future ...

Eigendecomposition of symmetric matrices is at the heart of many computer vision algorithms. However, the derivatives of the eigenvectors tend to be numerically unstable, whether using the SVD to compute them analytically or using the Power Iteration (PI) ...

Mathieu Salzmann

Chat with Graph Search

Source-Free Domain-Invariant Performance Prediction

Unsupervised Geometry-Aware Representation Learning for 3D Human Pose Estimation

HOISDF: Constraining 3D Hand-Object Pose Estimation with Global Signed Distance Fields: Processed data and trained models

Extending foundation models from weather to climate: challenges and promises

Learning to Reconstruct Texture-less Deformable Surfaces from a Single View

Visual Correspondences for Unsupervised Domain Adaptation on Electron Microscopy Images

GarNet++: Improving Fast and Accurate Static 3D Cloth Draping by Curvature Loss

Subseasonal-to-Seasonal strategies for the Earth System Foundation Model (ESFM)

Tiny satellites, big challenges: A feasibility study of machine vision pose estimation for PocketQubes during conjunctions

Match Normalization: Learning-Based Point Cloud Registration for 6D Object Pose Estimation in the Real World

Landmark Regularization: Ranking Guided Super-Net Training in Neural Architecture Search

On the Loss Landscape of Adversarial Training: Identifying Challenges and How to Overcome Them

Volumetric Transformer Networks

Local Non-Rigid Structure-from-Motion from Locally Diffeomorphic Mappings

Towards Stable and Efficient Adversarial Training against $l_1$ Bounded Adversarial Attacks

Linear-Covariance Loss for End-to-End Learning of 6D Pose Estimation

Open Books: Supporting the Comprehension of Photobooks as Sequential Art

Learning Monocular 3D Human Pose Estimation from Multi-view Images

A Stochastic Conditioning Scheme for Diverse Human Motion Prediction

Robust Differentiable SVD