arXiv subjects

Jean-Luc Meunier

Publications and source records attributed to Jean-Luc Meunier.

11 recordsLinked to original sources

arXiv2026

NAVER LABS Europe Submission to the Instruction-following 2026 Short Track

In this paper, we describe NAVER LABS Europe's submission to the instruction-following speech processing short track at IWSLT 2026. We participate again in the constrained setting, developing systems capable of jointly performing ASR, ST, and SQA from English speech into Chinese, Italian, and German. Building on our previous submission, ranked first in last year's short track, we update our multi-stage training pipeline by replacing the speech projector with SpeechMapper, a method for learning a speech-to-LLM embedding projector using only ASR data. In addition, we introduce a synthetic SQA dataset, fakACL, composed of artificially generated scientific presentations. This dataset is built by prompting the LLM backbone, segmenting the generated talks, and synthesizing speech with SeamlessM4T-large-v2. The combination of an improved speech projection mechanism and domain-specific synthetic data allows our model to outperform last year's best short-track system, while being considerably more compact and relying on a weaker LLM backbone. This year's results place our system tied for first place in the overall short track ranking.

cs.CL↗

arXiv2026

StarDrinks: An English and Korean Test Set for SLU Evaluation in a Drink Ordering Scenario

LLMs and speech assistants are increasingly used for task-oriented interactions, yet their evaluation often relies on controlled scenarios that fail to capture the variability and complexity of real user requests. Drink ordering, for example, involves diverse named entities, drink types, sizes, customizations, and brand-specific terminology, as well as spontaneous speech phenomena such as hesitations and self-corrections. To address this gap, we introduce StarDrinks, a test set in English and Korean containing speech utterances features, transcriptions, and annotated slots. Our dataset supports speech-to-slots SLU, transcription-to-slots NLU, and speech-to-transcription ASR evaluation, providing a realistic benchmark for model robustness and generalization in a linguistically rich, real-world task.

cs.CL↗

arXiv2025

Disentangled Object-Centric Image Representation for Robotic Manipulation

Learning robotic manipulation skills from vision is a promising approach for developing robotics applications that can generalize broadly to real-world scenarios. As such, many approaches to enable this vision have been explored with fruitful results. Particularly, object-centric representation methods have been shown to provide better inductive biases for skill learning, leading to improved performance and generalization. Nonetheless, we show that object-centric methods can struggle to learn simple manipulation skills in multi-object environments. Thus, we propose DOCIR, an object-centric framework that introduces a disentangled representation for objects of interest, obstacles, and robot embodiment. We show that this approach leads to state-of-the-art performance for learning pick and place skills from visual inputs in multi-object environments and generalizes at test time to changing objects of interest and distractors in the scene. Furthermore, we show its efficacy both in simulation and zero-shot transfer to the real world.

cs.CV↗

arXiv2024

Investigating the potential of Sparse Mixtures-of-Experts for multi-domain neural machine translation

We focus on multi-domain Neural Machine Translation, with the goal of developing efficient models which can handle data from various domains seen during training and are robust to domains unseen during training. We hypothesize that Sparse Mixture-of-Experts (SMoE) models are a good fit for this task, as they enable efficient model scaling, which helps to accommodate a variety of multi-domain data, and allow flexible sharing of parameters between domains, potentially enabling knowledge transfer between similar domains and limiting negative transfer. We conduct a series of experiments aimed at validating the utility of SMoE for the multi-domain scenario, and find that a straightforward width scaling of Transformer is a simpler and surprisingly more efficient approach in practice, and reaches the same performance level as SMoE. We also search for a better recipe for robustness of multi-domain systems, highlighting the importance of mixing-in a generic domain, i.e. Paracrawl, and introducing a simple technique, domain randomization.

cs.CL↗

arXiv2023

Dynamic Viscosity of Methane Hydrate Systems from Non-Einsteinian, Plasma-Functionalized Carbon Nanotube Nanofluids

The viscosity of oxygen-functionalized multi-walled carbon nanotube (O-MWCNT) nanofluids was measured for concentrations from 0.1 to 10 ppm under conditions of 0 to 30 MPag pressures and 0 to 10 C temperatures. The presence of O-MWCNTs did not affect the temperature dependence of viscosity but did reduce the effective viscosity of solution due to cumulative hydrogen bond-disrupting surface effects, which overcame internal drag forces. O-MWCNTs added a weak pressure dependence to the viscosity of solution because of their ability to align more with the flow direction as pressure increased. In the liquid to hydrate phase transition, the times to reach the maximum viscosity were faster in O-MWCNT systems compared to the pure water baseline. However, the presence of O-MWCNTs limited the conditions at which hydrates formed as increased nanoparticle collisions in those systems inhibited the formation of critical clusters of hydrate nuclei. The times to viscosity values most relevant to technological applications were minimally 28.02 % (200 mPa s) and 21.08 % (500 mPa s) slower than the baseline, both in the 1 ppm system, even though all systems were faster to the final viscosity. This was attributed to O-MWCNT entanglement, which resulted in a hydrate slurry occurring at lower viscosity values.

physics.app-ph↗

arXiv2023

Non-Einsteinian Viscosity Behavior in Plasma-Functionalized Graphene Nanoflake Nanofluids and their Effect on the Dynamic Viscosity of Methane Hydrate Systems

Water's viscosity dependence on pressure was also not affected by O-GNFs, except at 10 ppm, where the shuttle effect may have increased the presence of hydrophobic methane bubbles in the solution. Under high pressure, the relative viscosity of the system remained non-Einsteinian at all temperatures except 2C. This may have been because the density anomaly of water was shifted to a colder temperature as the hydrogen bonding network was weaker. The phase transition from liquid to hydrate was identical to that of pure water, indicating that the presence of different stages of growth was not affected by the presence of O-GNF. However, the times to reach a maximum viscosity were faster in O-GNF systems compared to pure water. This said, the hydrate formation limitations inherent to the measurement system were not overcome by the presence of O-GNFs. The times to application-relevant viscosity values were maximized in the 1 ppm system at 49.75 % (200 mPa.s) and 31.93 % (500 mPa.s) faster than the baseline. Therefore, the presence of O-GNFs allowed for shorter times to desired viscosities and at lower driving forces than the baseline, improving the viability of the hydrate technologies to which they can be added.

physics.app-ph↗

arXiv2022

LayoutXLM vs. GNN: An Empirical Evaluation of Relation Extraction for Documents

This paper investigates the Relation Extraction task in documents by benchmarking two different neural network models: a multi-modal language model (LayoutXLM) and a Graph Neural Network: Edge Convolution Network (ECN). For this benchmark, we use the XFUND dataset, released along with LayoutXLM. While both models reach similar results, they both exhibit very different characteristics. This raises the question on how to integrate various modalities in a neural network: by merging all modalities thanks to additional pretraining (LayoutXLM), or in a cascaded way (ECN). We conclude by discussing some methodological issues that must be considered for new datasets and task definition in the domain of Information Extraction with complex documents.

cs.CL↗

arXiv2019

Machine Translation of Restaurant Reviews: New Corpus for Domain Adaptation and Robustness

We share a French-English parallel corpus of Foursquare restaurant reviews (https://europe.naverlabs.com/research/natural-language-processing/machine-translation-of-restaurant-reviews), and define a new task to encourage research on Neural Machine Translation robustness and domain adaptation, in a real-world scenario where better-quality MT would be greatly beneficial. We discuss the challenges of such user-generated content, and train good baseline models that build upon the latest techniques for MT robustness. We also perform an extensive evaluation (automatic and human) that shows significant improvements over existing online systems. Finally, we propose task-specific metrics based on sentiment analysis or translation accuracy of domain-specific polysemous words.

cs.CL↗

arXiv2019

Comparing Machine Learning Approaches for Table Recognition in Historical Register Books

We present in this paper experiments on Table Recognition in hand-written registry books. We first explain how the problem of row and column detection is modeled, and then compare two Machine Learning approaches (Conditional Random Field and Graph Convolutional Network) for detecting these table elements. Evaluation was conducted on death records provided by the Archive of the Diocese of Passau. Both methods show similar results, a 89 F1 score, a quality which allows for Information Extraction. Software and dataset are open source/data.

cs.CV↗

arXiv2018

Bench-Marking Information Extraction in Semi-Structured Historical Handwritten Records

In this report, we present our findings from benchmarking experiments for information extraction on historical handwritten marriage records Esposalles from IEHHR - ICDAR 2017 robust reading competition. The information extraction is modeled as semantic labeling of the sequence across 2 set of labels. This can be achieved by sequentially or jointly applying handwritten text recognition (HTR) and named entity recognition (NER). We deploy a pipeline approach where first we use state-of-the-art HTR and use its output as input for NER. We show that given low resource setup and simple structure of the records, high performance of HTR ensures overall high performance. We explore the various configurations of conditional random fields and neural networks to benchmark NER on given certain noisy input. The best model on 10-fold cross-validation as well as blind test data uses n-gram features with bidirectional long short-term memory.

cs.CV↗

arXiv2017

Joint Structured Learning and Predictions under Logical Constraints in Conditional Random Fields

This paper is concerned with structured machine learning, in a supervised machine learning context. It discusses how to make joint structured learning on interdependent objects of different nature, as well as how to enforce logical con-straints when predicting labels. We explain how this need arose in a Document Understanding task. We then discuss a general extension to Conditional Random Field (CRF) for this purpose and present the contributed open source implementation on top of the open source PyStruct library. We evaluate its performance on a publicly available dataset.

stat.ML↗