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

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

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

Mathieu Salzmann, Jean-Paul Richard Kneib, 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

MotionMap: Representing Multimodality in Human Pose Forecasting

Mathieu Salzmann, Alexandre Massoud Alahi, Saeed Saadatnejad, Megh Hiren Shukla

Human pose forecasting is inherently multimodal since multiple futures exist for an observed pose sequence. However, evaluating multimodality is challenging since the task is illposed. Therefore, we first propose an alternative paradigm to make the task we ...

2025

Dvmnet: Computing Relative Pose for Unseen Objects Beyond Hypotheses

Mathieu Salzmann, Zheng Dang

Determining the relative pose of an object between two images is pivotal to the success of generalizable object pose estimation. Existing approaches typically approximate the continuous pose representation with a large number of discrete pose hypotheses, w ...

IEEE2024

NOPE: Novel Object Pose Estimation from a Single Image

Mathieu Salzmann

The practicality of 3D object pose estimation remains limited for many applications due to the need for prior knowledge of a 3D model and a training period for new objects. To address this limitation, we propose an approach that takes a single image of a n ...

IEEE2024

Synthetic data (Part 1) for "HOISDF: Constraining 3D Hand-Object Pose Estimation with Global Signed Distance Fields"

Mathieu Salzmann, Alexander Mathis

Here we provide the data of our article "HOISDF: Constraining 3D Hand-Object Pose Estimation with Global Signed Distance Fields". It contains the preprocessed SDF samples. Meanwhile, we also include rendered data for HO3Dv2 here. The overall structure of ...

Zenodo2024

Synthetic data (Part 2) for "HOISDF: Constraining 3D Hand-Object Pose Estimation with Global Signed Distance Fields"

Mathieu Salzmann, Alexander Mathis

Here we provide the data of our article "HOISDF: Constraining 3D Hand-Object Pose Estimation with Global Signed Distance Fields". It contains the rendered images and the segmentation masks that we use to train our model on HO3Dv2 dataset. The overall str ...

Zenodo2024

Mitigating Object Dependencies: Improving Point Cloud Self-Supervised Learning through Object Exchange

Mathieu Salzmann, Sabine Süsstrunk, Tong Zhang, Yi Wu

In the realm of point cloud scene understanding, particularly in indoor scenes, objects are arranged following human habits, resulting in objects of certain semantics being closely positioned and displaying notable inter-object correlations. This can creat ...

2024

Using Motion Cues to Supervise Single-Frame Body Pose and Shape Estimation in Low Data Regimes

Pascal Fua, Mathieu Salzmann, Andrey Davydov

When enough annotated training data is available, supervised deep-learning. algorithms excel at estimating human body pose and shape using a single camera. The effects of too little such data being available can be mitigated by using other information sour ...

2024

A Closed-Form Pairwise Solution to Local Non-Rigid Structure-from-Motion

Pascal Fua, Mathieu Salzmann, Shaifali Parashar, Yuxuan Long

2024

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

Mathieu Salzmann, Zheng Dang

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 ...

Ieee Computer Soc2024

GigaPose: Fast and Robust Novel Object Pose Estimation via One Correspondence

Mathieu Salzmann

We present GigaPose, a fast, robust, and accurate method for CAD-based novel object pose estimation in RGB images. GigaPose first leverages discriminative "templates", rendered images of the CAD models, to recover the out-of-plane rotation and then uses pa ...

IEEE2024