arXiv subjects

Thibaut Durand

Publications and source records attributed to Thibaut Durand.

12 recordsLinked to original sources

arXiv2026

Contact mechanics and friction of soft materials: an apparatus combining multi-axes dynamical actuation/measurement and in situ/in operando visualisation

The mechanics and friction of contact interfaces involving soft materials like gel, rubber or human skin are of both fundamental and applied interest. Recent insights have been made into this field thanks to in situ observations of the contact interface. However, current soft-material-oriented tribometers enable only few degrees of freedom for the actuation of the contact, far from covering the richness of the loading conditions relevant to real tribological contacts. Here, we introduce an apparatus dedicated to the study of the contact mechanics and friction of soft materials which, in addition to in situ / in operando optical monitoring of the interface, enables simultaneous actuation along five degrees of freedom: three translations and two rotations. While all three translations feature large velocity/large stroke motion, the one responsible for normal contact loading can also apply high frequency/small amplitude vibrations. The contact's dynamical response is monitored using both a 6-axes force/torque sensor and a 6-axes displacement/rotation sensor. We first describe the structure of the apparatus, its implementation, alignment, calibration and resolutions. We then illustrate its capabilities through a series of experiments on elastomer contacts. Our apparatus will be useful to investigate the mechanics of a wide range of soft interfaces submitted to rich, tribology-relevant kinematic or dynamic stimuli.

physics.class-ph↗

arXiv2026

FairNVT: Fair Classification via Noise Injection in Vision Transformers

This paper presents FairNVT, a lightweight debiasing framework for pretrained transformer-based encoders that improves prediction fairness while preserving task performance. FairNVT is motivated by the intuition that reducing sensitive-attribute information in the representation used by the downstream classifier can facilitate fairer predictions. Our approach learns task-relevant and sensitive embeddings via lightweight adapters, applies calibrated Gaussian noise to the sensitive embedding, and fuses it with the task representation. Together with orthogonality constraints and fairness regularization, these components jointly reduce sensitive-attribute leakage in the learned embeddings and encourage fairer downstream predictions. Across three datasets spanning vision and language, FairNVT reduces sensitive-attribute attacker accuracy, improves fairness metrics such as demographic parity difference and equalized odds, and maintains competitive task performance.

cs.CV↗

arXiv2026

ReDiTT: Retrieval Augmented Conditional Diffusion Transformers for Asynchronous Time Series

We present a diffusion based model for asynchronous time series prediction, where the goal is to predict the next inter event time and event type. To address the inherent uncertainty of future events, we introduce ReDiTT, a retrieval augmented conditional diffusion transformer that operates in latent space. ReDiTT retrieves structurally similar latent sequences from a memory bank during both training and inference and incorporates them as reference conditions through cross attention. This retrieval based conditioning allows the model to attend to relevant temporal dynamics and provides global structural guidance for generation. As a result, ReDiTT stabilizes long horizon forecasting and improves sample diversity. Experiments on seven real world datasets demonstrate state of the art performance on next event prediction and long horizon forecasting. Our code is available at https://github.com/BorealisAI/ReDiTT.

cs.LG↗

arXiv2026

An Attention-based Model for Robust Forecasting with Missing Modality

Learning with missing modalities is a fundamental challenge in multimodal robot learning, as real-world robotic systems often operate in environments with incomplete sensor data. Attention-based models are appealing for processing multimodal data because they can handle multiple modalities with a single backbone network. However, most multimodal models assume that all modalities are available during both training and inference, limiting their applicability in robotic perception and decision-making. In this paper, we introduce a multimodal model designed to handle missing modalities during both training and inference. The model is formulated as a conditional variational autoencoder (CVAE) and incorporates a transformer-based architecture that leverages attention mechanisms to learn a unified, fixed-dimensional representation, even when some modalities are missing. We show that our proposed model can be trained with missing modalities while approximating a robust representation of all modalities. We evaluate our approach on five multimodal datasets across two robot learning tasks: human trajectory prediction and robot manipulation forecasting. Experimental results demonstrate that our model effectively learns from incomplete data and is superior to prior multimodal fusion approaches.

cs.RO↗

arXiv2025

LAST SToP For Modeling Asynchronous Time Series

We present a novel prompt design for Large Language Models (LLMs) tailored to Asynchronous Time Series. Unlike regular time series, which assume values at evenly spaced time points, asynchronous time series consist of timestamped events occurring at irregular intervals, each described in natural language. Our approach effectively utilizes the rich natural language of event descriptions, allowing LLMs to benefit from their broad world knowledge for reasoning across different domains and tasks. This allows us to extend the scope of asynchronous time series analysis beyond forecasting to include tasks like anomaly detection and data imputation. We further introduce Stochastic Soft Prompting, a novel prompt-tuning mechanism that significantly improves model performance, outperforming existing fine-tuning methods such as QLoRA. Through extensive experiments on real world datasets, we demonstrate that our approach achieves state-of-the-art performance across different tasks and datasets.

cs.LG↗

arXiv2022

Training a Vision Transformer from scratch in less than 24 hours with 1 GPU

Transformers have become central to recent advances in computer vision. However, training a vision Transformer (ViT) model from scratch can be resource intensive and time consuming. In this paper, we aim to explore approaches to reduce the training costs of ViT models. We introduce some algorithmic improvements to enable training a ViT model from scratch with limited hardware (1 GPU) and time (24 hours) resources. First, we propose an efficient approach to add locality to the ViT architecture. Second, we develop a new image size curriculum learning strategy, which allows to reduce the number of patches extracted from each image at the beginning of the training. Finally, we propose a new variant of the popular ImageNet1k benchmark by adding hardware and time constraints. We evaluate our contributions on this benchmark, and show they can significantly improve performances given the proposed training budget. We will share the code in https://github.com/BorealisAI/efficient-vit-training.

cs.CV↗

arXiv2021

LayoutVAE: Stochastic Scene Layout Generation From a Label Set

Recently there is an increasing interest in scene generation within the research community. However, models used for generating scene layouts from textual description largely ignore plausible visual variations within the structure dictated by the text. We propose LayoutVAE, a variational autoencoder based framework for generating stochastic scene layouts. LayoutVAE is a versatile modeling framework that allows for generating full image layouts given a label set, or per label layouts for an existing image given a new label. In addition, it is also capable of detecting unusual layouts, potentially providing a way to evaluate layout generation problem. Extensive experiments on MNIST-Layouts and challenging COCO 2017 Panoptic dataset verifies the effectiveness of our proposed framework.

cs.CV↗

arXiv2021

Variational Selective Autoencoder: Learning from Partially-Observed Heterogeneous Data

Learning from heterogeneous data poses challenges such as combining data from various sources and of different types. Meanwhile, heterogeneous data are often associated with missingness in real-world applications due to heterogeneity and noise of input sources. In this work, we propose the variational selective autoencoder (VSAE), a general framework to learn representations from partially-observed heterogeneous data. VSAE learns the latent dependencies in heterogeneous data by modeling the joint distribution of observed data, unobserved data, and the imputation mask which represents how the data are missing. It results in a unified model for various downstream tasks including data generation and imputation. Evaluation on both low-dimensional and high-dimensional heterogeneous datasets for these two tasks shows improvement over state-of-the-art models.

cs.LG↗

arXiv2021

Design and characterization of an instrumented slider aimed atmeasuring local micro-impact forces between dry rough solids

Sliding motion between two rough solids under light normal loading involves myriad micro-impacts between antagonist micro-asperities. Those micro-impacts are at the origin of many emerging macroscopic phenomena, including the friction force, the slider's vibrations and the noise radiated in the surroundings. However, the individual properties of the micro-impacts (e.g. maximum force, position along the interface, duration) are essentially elusive to measurement. Here, we introduce an instrumented slider aimed at measuring the position and the normal component of the micro-impact forces during sliding against a rough track. It is based on an array of piezoelectric sensors, each placed under a single model asperity. Its dynamical characteristics are established experimentally and compared to a finite elements model. We then validate its measurement capabilities by using it against a track bearing simple, well-defined topographical features. The measurements are interpreted thanks to a simple multi-asperity contact model. Our slider is expected to be useful in future studies to provide local insights into a variety of tribological questions involving dry rough sliding interfaces.

physics.class-ph↗

arXiv2019

Point Process Flows

Event sequences can be modeled by temporal point processes (TPPs) to capture their asynchronous and probabilistic nature. We propose an intensity-free framework that directly models the point process distribution by utilizing normalizing flows. This approach is capable of capturing highly complex temporal distributions and does not rely on restrictive parametric forms. Comparisons with state-of-the-art baseline models on both synthetic and challenging real-life datasets show that the proposed framework is effective at modeling the stochasticity of discrete event sequences.

cs.LG↗

arXiv2019

A Variational Auto-Encoder Model for Stochastic Point Processes

We propose a novel probabilistic generative model for action sequences. The model is termed the Action Point Process VAE (APP-VAE), a variational auto-encoder that can capture the distribution over the times and categories of action sequences. Modeling the variety of possible action sequences is a challenge, which we show can be addressed via the APP-VAE's use of latent representations and non-linear functions to parameterize distributions over which event is likely to occur next in a sequence and at what time. We empirically validate the efficacy of APP-VAE for modeling action sequences on the MultiTHUMOS and Breakfast datasets.

cs.CV↗

arXiv2019

Learning a Deep ConvNet for Multi-label Classification with Partial Labels

Deep ConvNets have shown great performance for single-label image classification (e.g. ImageNet), but it is necessary to move beyond the single-label classification task because pictures of everyday life are inherently multi-label. Multi-label classification is a more difficult task than single-label classification because both the input images and output label spaces are more complex. Furthermore, collecting clean multi-label annotations is more difficult to scale-up than single-label annotations. To reduce the annotation cost, we propose to train a model with partial labels i.e. only some labels are known per image. We first empirically compare different labeling strategies to show the potential for using partial labels on multi-label datasets. Then to learn with partial labels, we introduce a new classification loss that exploits the proportion of known labels per example. Our approach allows the use of the same training settings as when learning with all the annotations. We further explore several curriculum learning based strategies to predict missing labels. Experiments are performed on three large-scale multi-label datasets: MS COCO, NUS-WIDE and Open Images.

cs.CV↗