Complex data, machine learning and representations

Our research team focus on the problems raised by large-scale data management, with a strong orientation towards data whose structure, explicit or not, is complex and requires specific techniques of approximation, extraction and search. The type of data we are dealing with include images, videos, audio or musical documents, satellite imagery and data from multi-spectral sensors. We investigate techniques of statistical machine learning (with a specific focus on deep learning) to extract information, build efficient access techniques and propose new methods of data management based directly on content (as opposed to metadata describing this content). At the moment there are two research axis/directions in our team, described below.

Axis 1. Large image and video databases
We live today in a context characterized by an explosive growth in the production of digital content, doubled by a revolution in digital storage making it possible to keep and easily access large quantities of digital data, beyond the timeline for which it has been initially collected. On the other hand, the rapid development of digital transmission technologies makes possible the distributed distribution and remote sharing of large volumes of such a digital contents. We focus on the structuring, from visual content, of large image and video databases, as well as the search by content in such databases. Our recent work focus on deep learning for the detection of visual patterns and for semantic segmentation of images, the goal being the semantic analysis of scenes taking into account structural and global-local relationships between image components. These approaches also apply very well to data of a different nature, such as musical data, which combine structures at different scales and are generally characterized by a relatively small number of structural items labeled by class.

Axis 2. Music computing and music information retrieval
This axis of research aims to investigate the production of models of musical languages, characterizing homogeneous corpora of music available in symbolic form (scores). Our perspective is to enrich a statistical approach based on explicit data (notes) by a knowledge extraction process identifying the elements of musical language implicitly present in the notation: segmentation in phrases, presence and use of patterns, management of dissonances, cadences, instrumentation and texture. Another direction of research is the development of automatic transcription techniques, conversion of a musical performance to a score in traditional notation by a priori score models (independent of the performance to be transcribed), representing the language of possible musical notations. These techniques can be seen as language models, and are essential components of machine translation or pattern recognition procedures for music data (by analogy with natural language processing).

Publications

2021

Articles de revue

  1. Castillo-Navarro, J.; Le Saux, B.; Boulch, A.; Audebert, N. and Lefèvre, S. Semi-Supervised Semantic Segmentation in Earth Observation: The MiniFrance suite, dataset analysis and multi-task network study. In Machine Learning, 2021. doi  www 

Articles de conférence

  1. Foscarin, F.; Audebert, N. and Fournier-S'niehotta, R. PKSpell: Data-Driven Pitch Spelling and Key Signature Estimation. In International Society for Music Information Retrieval Conference (ISMIR), Online, India, 2021. www 
  1. Le Guen, V.; Yin, Y.; Dona, J.; Ayed, I.; de Bézenac, E.; Thome, N. and Gallinari, P. Augmenting physical models with deep networks for complex dynamics forecasting. In Ninth International Conference on Learning Representations ICLR 2021, Vienne (virtual), Austria, 2021. www 
  1. Dang, C.; Randrianarivo, H.; Fournier-S'niehotta, R. and Audebert, N. Web Image Context Extraction with Graph Neural Networks and Sentence Embeddings on the DOM tree. In GEM: Graph Embedding and Mining - ECML/PKDD Workshops, Bilbao, Spain, 2021. www 

2020

Articles de revue

  1. Besedin, A.; Blanchart, P.; Crucianu, M. and Ferecatu, M. Deep online classification using pseudo-generative models. In Computer Vision and Image Understanding, 201: 103048, 2020. doi  www 
  1. Dupuis, D.; Du Mouza, C.; Travers, N. and Chareyron, G. Real-Time Influence Maximization in a RTB Setting. In Data Science and Engineering, 5 (3): 224-239, 2020. doi  www 
  1. Rosmorduc, S. Automated~Transliteration~of Late Egyptian Using Neural Networks: An Experiment in ``Deep Learning''. In Lingua Aegyptia - Journal of Egyptian Language Studies, 28: 233-257, 2020. doi  www 
  1. rambour, c.; Audebert, N.; Koeniguer, E.; Le Saux, B.; Crucianu, M. and Datcu, M. Flood detection in time series of optical and sar images. In International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLIII-B2-2020: 1343-1346, 2020. doi  www 

Articles de conférence

  1. Foscarin, F.; Mcleod, A.; Rigaux, P.; Jacquemard, F. and Sakai, M. ASAP: a dataset of aligned scores and performances for piano transcription. In ISMIR 2020 - 21st International Society for Music Information Retrieval, Montreal / Virtual, Canada, 2020. www 
  1. Le Cacheux, Y.; Le Borgne, H. and Crucianu, M. Using Sentences as Semantic Representations in Large Scale Zero-Shot Learning. In ECCV 2020 workshop Transferring and adapting source knowledge in computer vision (TASK-CV), Glasgow, United Kingdom, 2020. www 
  1. Le Guen, V. and Thome, N. Probabilistic Time Series Forecasting with Structured Shape and Temporal Diversity. In NeurIPS 2020, Vancouver, Canada, 2020. www 
  1. Le Guen, V. and Thome, N. A Deep Physical Model for Solar Irradiance Forecasting with Fisheye Images. In CVPR OmniCV worshop 2020, Seattle, United States, 2020. doi  www 
  1. Le Guen, V. and Thome, N. Disentangling Physical Dynamics from Unknown Factors for Unsupervised Video Prediction. In Computer Vision and Pattern Recognition 2020 (CVPR), Seattle, United States, 2020. doi  www 
  1. Dubucq, D.; Audebert, N.; Achard, V.; Alakian, A.; Fabre, S.; Credoz, A.; Deliot, P. and Le Saux, B. A real-world hyperspectral image processing workflow for vegetatotion stress and hydrocarbon indirect detection. In XXIV ISPRS Congress, Nice, France, The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLIII-B3-2020, 2020. doi  www 
  1. Mali, J.; Atigui, F.; Azough, A. and Travers, N. ModelDrivenGuide: An Approach for Implementing NoSQL Schemas. In International Conference, DEXA 2020, pages 141-151, Springer, Bratislava, Slovakia, DEXA 2020: Database and Expert Systems Applications , 2020. doi  www 
  1. Rolland, J-F. c.; Castel, F.; Haugommard, A.; Aubrun, M.; Yao, W.; Dumitru, C. O.; Datcu, M.; Bylicki, M.; Tran, B-H.; Aussenac-Gilles, N.; Comparot, C. and Trojahn dos Santos, C. CANDELA: A Cloud Platform for Copernicus Earth Observation Data Analytics. In IEEE International Geoscience & Remote Sensing Symposium - IGARSS 2020, IEEE, Waikoloa, Hawaii, United States, 2020. www 

2019

Articles de revue

  1. Viard, T. and Fournier-S'niehotta, R. Augmenting content-based rating prediction with link stream features. In Computer Networks, 150: 127-133, 2019. doi  www 
  1. Lajaunie, C.; Renard, D.; Quentin, A.; Le Guen, V. and Caffari, Y. A non-homogeneous model for kriging dosimetric data. In Mathematical Geosciences, 52 (7): 847-863, 2019. doi  www 

Articles de conférence

  1. Viard, T. and Fournier-S'niehotta, R. Encoding temporal and structural information in machine learning models for recommendation. In LEG @ ECML-PKDD 2019, W"urzburg, Germany, 2019. www 
  1. Dupuis, D.; Du Mouza, C.; Travers, N. and Chareyron, G. RTIM: a Real-Time Influence Maximization Strategy. In Web Information Systems Engineering -- WISE 2019, Hong-Kong, China, 2019. doi  www 
  1. Grossetti, Q.; Du Mouza, C. and Travers, N. Community-based Recommendations on Twitter: Avoiding The Filter Bubble. In Web Information Systems Engineering -- WISE 2019, Hong-Kong, China, 2019. doi  www 
  1. Rigaux, P. and Travers, N. Scalable Searching and Ranking for Melodic Pattern Queries. In Intl. Conf. of the International Society for Music Information Retrieval (ISMIR), Delft, Netherlands, 2019. www 
  1. Foscarin, F.; Jacquemard, F. and Rigaux, P. Modeling and Learning Rhythm Structure. In Sound and Music Computing Conference (SMC), Malaga, Spain, 2019. www 
  1. Foscarin, F.; Fournier-S'niehotta, R. and Jacquemard, F. A diff procedure for music score files. In 6th International Conference on Digital Libraries for Musicology (DLfM), pages 7, ACM, The Hague, Netherlands, 2019. www 
  1. Foscarin, F.; Jacquemard, F.; Rigaux, P. and Sakai, M. A Parse-based Framework for Coupled Rhythm Quantization and Score Structuring. In MCM 2019 - Mathematics and Computation in Music, Springer, Madrid, Spain, Proceedings of the Seventh International Conference on Mathematics and Computation in Music (MCM 2019) Lecture Notes in Computer Science, 2019. doi  www 
  1. Le Cacheux, Y.; Le Borgne, H. and Crucianu, M. Modeling Inter and Intra-Class Relations in the Triplet Loss for Zero-Shot Learning. In IEEE International Conference on Computer Vision, IEEE, Séoul, South Korea, 2019. doi  www 

2018

Articles de revue

  1. Raftopoulos, K.; Kollias, S.; Sourlas, D. and Ferecatu, M. On the Beneficial Effect of Noise in Vertex Localization. In International Journal of Computer Vision, 126 (1): 111-139, 2018. doi  www 
  1. Fournier-S'niehotta, R.; Rigaux, P. and Travers, N. Modeling Music as Synchronized Time Series: Application to Music Score Collections. In Information Systems, 73: 35-49, 2018. doi  www 

Articles de conférence

  1. Besedin, A.; Blanchart, P.; Crucianu, M. and Ferecatu, M. Deep Online Storage-Free Learning on Unordered Image Streams. In ECML PKDD: Joint European Conference on Machine Learning and Knowledge Discovery in Databases, pages 103-112, Dublin, Ireland, Communications in Computer and Information Science - ECML PKDD 2018 Workshops 967, 2018. doi  www 
  1. Grossetti, Q.; Constantin, C.; Du Mouza, C. and Travers, N. An Homophily-based Approach for Fast Post Recommendation in Microblogging Systems. In 21st International Conference on Extending Database Technology (EDBT 2018), pages 229-240, Vienne, Austria, 2018. doi  www 
  1. Petit, O.; Thome, N.; Charnoz, A.; Hostettler, A. and Soler, L. Handling Missing Annotations for Semantic Segmentation with Deep ConvNets. In MICCAI workshop DLMIA, Grenade, Spain, 2018. doi  www 
  1. Foscarin, F.; Fiala, D.; Jacquemard, F.; Rigaux, P. and Thion, V. Gioqoso, an online Quality Assessment Tool for Music Notation. In 4th International Conference on Technologies for Music Notation and Representation (TENOR'18), Concordia University, Montreal, Canada, Proceedings of the International Conference on Technologies for Music Notation and Representation -- TENOR'18 , 2018. www 

2017

Articles de conférence

  1. Besedin, A.; Blanchart, P.; Crucianu, M. and Ferecatu, M. Evolutive deep models for online learning on data streams with no storage. In ECML/PKDD 2017 Workshop on Large-scale Learning from Data Streams in Evolving Environments, Skopje, Macedonia, 2017. www 
  1. Cherfi, S. S-s.; Guillotel, C.; Hamdi, F. c.; Rigaux, P. and Travers, N. Ontology-Based Annotation of Music Scores. In Knowledge Capture Conference, pages 1-4, ACM Press, Austin, France, 2017. doi  www 
  1. Si-Said Cherfi, S.; Hamdi, F. c.; Rigaux, P.; Thion, V. and Travers, N. Formalizing quality rules on music notation. An ontology-based approach. In International Conference on Technologies for Music Notation and Representation - TENOR'17, Coruna, Spain, 2017. www 
  1. Rigaux, P. and Thion, V. Quality Awareness over Graph Pattern Queries. In Proceedings of the International Database Engineering & Applications Symposium (IDEAS), Bristol, United Kingdom, 2017. doi  www 

Communications

Publications

Softwares and patents

Publications

Ongoing projects

    Past projects

      Top