Publications by Mathieu Salzmann

3D Single-Object Tracking in Point Clouds with High Temporal Variation

The high temporal variation of the point clouds is the key challenge of 3D single-object tracking (3D SOT). Existing approaches rely on the assumption that the shape variation of the point clouds and the motion of the objects across neighboring frames are ...

Springer Science and Business Media Deutschland GmbH2025

6Img-to-3D: Few-Image Large-Scale Outdoor Novel View Synthesis

Mathieu Salzmann

Current 3D reconstruction techniques struggle to infer unbounded scenes from a few images faithfully. Most existing methods have high computational demands, require detailed pose information, and cannot reconstruct occluded regions reliably. We introduce 6 ...

IEEE2025

Towards Self-Supervised Covariance Estimation in Deep Heteroscedastic Regression

Mathieu Salzmann, Alexandre Massoud Alahi, Megh Hiren Shukla

Deep heteroscedastic regression models the mean and covariance of the target distribution through neural networks. The challenge arises from heteroscedasticity, which implies that the covariance is sample dependent and is often unknown. Consequently, recen ...

ICLR2025

A method for asteroid detection using convolutional neural networks on VST images

Jean-Paul Richard Kneib, Mathieu Salzmann, Stephan Hellmich, Elisabeth Andréa Cécile Rachith

Context. The study of asteroids, particularly near-Earth asteroids, is key to gaining insights into our Solar System and can help prevent dangerous collisions. Beyond finding new objects, additional observations of known asteroids will improve our knowledg ...

2025

Source-Free Domain-Invariant Performance Prediction

Mathieu Salzmann

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 ...

Springer Nature2025

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

Mathieu Salzmann

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 ...

2025

Deep Learning to Improve the Discovery of Near-earth Asteroids in the Zwicky Transient Facility

Jean-Paul Richard Kneib, Mathieu Salzmann, Belén Yu Irureta-Goyena Chang

We present a novel pipeline that uses a convolutional neural network (CNN) to improve the detection capability of near-Earth asteroids (NEAs) in the context of planetary defense. Our work aims to minimize the dependency on human intervention of the current ...

IOP Publishing Ltd2025

Joint Compression of Multi-model Systems for Edge Devices

Mathieu Salzmann

Deploying deep learning models for computer vision (CV) often requires a careful trade-off between performance, efficiency, and cost. In the context of edge inference of machine learning systems, this trade-off translates into a challenge, as computing and ...

Springer Science and Business Media Deutschland GmbH2025

Temporally-Consistent Surface Reconstruction Using Metrically-Consistent Atlases

Pascal Fua, Mathieu Salzmann, Shaifali Parashar, Jan Bednarík

We propose a method for unsupervised reconstruction of a temporally-consistent sequence of surfaces from a sequence of time-evolving point clouds. It yields dense and semantically meaningful correspondences between frames. We represent the reconstructed su ...

2025

SINDER: Repairing the Singular Defects of DINOv2

Mathieu Salzmann, Haoqi Wang

Vision Transformer models trained on large-scale datasets, although effective, often exhibit artifacts in the patch token they extract. While such defects can be alleviated by re-training the entire model with additional classification tokens, the underlyi ...

Springer Nature2025