Searcharxiv⌕ Search

arXiv subjects

Francis Rousseaux

Publications and source records attributed to Francis Rousseaux.

7 recordsLinked to original sources
arXiv2023

Named Entity Recognition for Monitoring Plant Health Threats in Tweets: a ChouBERT Approach

An important application scenario of precision agriculture is detecting and measuring crop health threats using sensors and data analysis techniques. However, the textual data are still under-explored among the existing solutions due to the lack of labelled data and fine-grained semantic resources. Recent research suggests that the increasing connectivity of farmers and the emergence of online farming communities make social media like Twitter a participatory platform for detecting unfamiliar plant health events if we can extract essential information from unstructured textual data. ChouBERT is a French pre-trained language model that can identify Tweets concerning observations of plant health issues with generalizability on unseen natural hazards. This paper tackles the lack of labelled data by further studying ChouBERT's know-how on token-level annotation tasks over small labeled sets.

cs.CL↗
arXiv2023

Combining pre-trained Vision Transformers and CIDER for Out Of Domain Detection

Out-of-domain (OOD) detection is a crucial component in industrial applications as it helps identify when a model encounters inputs that are outside the training distribution. Most industrial pipelines rely on pre-trained models for downstream tasks such as CNN or Vision Transformers. This paper investigates the performance of those models on the task of out-of-domain detection. Our experiments demonstrate that pre-trained transformers models achieve higher detection performance out of the box. Furthermore, we show that pre-trained ViT and CNNs can be combined with refinement methods such as CIDER to improve their OOD detection performance even more. Our results suggest that transformers are a promising approach for OOD detection and set a stronger baseline for this task in many contexts

cs.CV↗
arXiv2021

Fine-tuning BERT-based models for Plant Health Bulletin Classification

In the era of digitization, different actors in agriculture produce numerous data. Such data contains already latent historical knowledge in the domain. This knowledge enables us to precisely study natural hazards within global or local aspects, and then improve the risk prevention tasks and augment the yield, which helps to tackle the challenge of growing population and changing alimentary habits. In particular, French Plants Health Bulletins (BSV, for its name in French Bulletin de Sant{é} du V{é}g{é}tal) give information about the development stages of phytosanitary risks in agricultural production. However, they are written in natural language, thus, machines and human cannot exploit them as efficiently as it could be. Natural language processing (NLP) technologies aim to automatically process and analyze large amounts of natural language data. Since the 2010s, with the increases in computational power and parallelization, representation learning and deep learning methods became widespread in NLP. Recent advancements Bidirectional Encoder Representations from Transformers (BERT) inspire us to rethink of knowledge representation and natural language understanding in plant health management domain. The goal in this work is to propose a BERT-based approach to automatically classify the BSV to make their data easily indexable. We sampled 200 BSV to finetune the pretrained BERT language models and classify them as pest or/and disease and we show preliminary results.

cs.CL↗
arXiv2009

Music-ripping: des pratiques qui provoquent la musicologie

Out of the scope of the usual positions of computing in the field of music and musicology, one notices the emergence of human-computer systems that do exist by breaking off. Though these singular systems take effect in the usual fields of expansion of music, they do not make any systematic reference to known musicological categories. On the contrary, they make possible experiments that open uses where listening, composition and musical transmission get merged in a gesture sometimes named as ?music-ripping?. We will show in which way the music-ripping practices provoke traditional musicology, whose canonical categories happen to be ineffectual to explain here. To achieve that purpose, we shall need: - to make explicit a minimal set of categories that is sufficient to underlie the usual models of computer assisted music;- to do the same for human-computer systems (anti-musicological?) that disturb us; - to examine the possibility conditions of reduction of the second set to the first; - to conclude on the nature of music-ripping.

cs.MM↗
arXiv2009

Interagir avec un contenu opératique : le projet d'opéra virtuel interactif Virtualis

In this article, we present the interactive opera project on CD-ROM Virtualis. This project includes a scientific dimension as well as artistic. It gave us the opportunity to design a model of the opera performance using formalisms from organization sciences. Moreover, our investigation on interactions between a user and opera contents led us to use models of relationships between entities based on physical forces, where the user is in a way absent. We detail some aspects of a reading but also writing environment on artistic complex contents between text, music and graphics.

cs.HC↗