Skip to Main content

Researcher identifiers

Social networks

Export Publications

Export the displayed publications:
Number of documents

22

Sahar Ghannay

Associate professort at LIMSI, CNRS, Université Paris-Saclay

Team: ILES

Email: sahar.ghannay@limsi.fr

Website: https://saharghannay.github.io

 

Short Bio

Sahar Ghannay is an associate professor at Université Paris-Saclay, in the CNRS, LISN research center, since September 2018.

She received a PhD in Computer Science from Le Mans University on Septembre 2017. Her thesis work is part of the ANR VERA (AdVanced ERror Analysis for speech recognition) project. During her PhD, she spent a few months as @ visiting researcher at Apple within the Siri Speech team.

As a postdoctoral researcher at LIUM, she worked on neural end-to-end systems for the detection of named entities, speech understanding, as part of the Chist-Era M2CR (Multimodal Multilingual Continuous Representation for Human Language Understanding) project.

Her main research interests are continuous representations learning and their application to natural language processing and speech recognition tasks, semantic information extraction form spoken and writen language and dialog system.

 

CV

Education

  • PHD in computer science, at LIUM, Le Mans université, 2017
  • MS in computer science, Le Mans université, 2013
  • BS in computer science,  Le Mans université and université de sfax, 2011

Work Experience 

  • 2018 - now: Associate Professort at LISN, CNRS, Université Paris-Saclay
  • 2017-2018: Post-doc at LIUM, Le Mans université
  • 2017 (4 moths): Internship at Apple within the Siri Speech team at Cupertino
  • April 2013-Sept. 2014: research engineer

Journal articles1 document

  • Sahar Ghannay, Yannick Estève, Nathalie Camelin. A study of continuous space word and sentence representations applied to ASR error detection. Speech Communication, Elsevier : North-Holland, 2020. ⟨hal-02501943⟩

Conference papers21 documents

  • Juan Manuel Coria, Hervé Bredin, Sahar Ghannay, Sophie Rosset. OVERLAP-AWARE LOW-LATENCY ONLINE SPEAKER DIARIZATION BASED ON END-TO-END LOCAL SEGMENTATION. IEEE Automatic Speech Recognition and Unserstanding Workshop, Dec 2021, Cartagena, Colombia. ⟨hal-03375330⟩
  • Sahar Ghannay, Christophe Servan, Sophie Rosset. Neural Networks approaches focused on French Spoken Language Understanding: application to the MEDIA Evaluation Task. In Proceedings of The 28th International Conference on Computational Linguistics (COLING’2020), 2020, Dec 2020, Barcelona (online), Spain. ⟨hal-03007482⟩
  • Mathilde Veron, Anselmo Peã±as, Guillermo Echegoyen, Somnath Banerjee, Sahar Ghannay, et al.. A Cooking Knowledge Graph and Benchmark for Question Answering Evaluation in Lifelong Learning Scenarios. International Conference on Applications of Natural Language to Information Systems, Elisabeth Métais and Farid Meziane and Helmut Horacek and Philipp Cimiano, Jun 2020, Saarbrücken, Germany. ⟨hal-03006228⟩
  • Juan Manuel Coria, Hervé Bredin, Sahar Ghannay, Sophie Rosset. A Comparison of Metric Learning Loss Functions for End-To-End Speaker Verification. International Conference on Statistical Language and Speech Processing, Oct 2020, Cardiff, United Kingdom. pp.137-148, ⟨10.1007/978-3-030-59430-5_11⟩. ⟨hal-02989334⟩
  • Sahar Ghannay, Antoine Neuraz, Sophie Rosset. What is best for Spoken Language Understanding: Small but Task-dependant Embeddings or Huge but Out-of-domain Embeddings?. IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), May 2020, Barcelone, Spain. ⟨hal-02503694⟩
  • Antoine Caubrière, Sahar Ghannay, Natalia Tomashenko, Renato de Mori, Antoine Laurent, et al.. Error analysis applied to end-to end spoken language understanding. 45th International Conference on Acoustics, Speech, and Signal Processing (ICASSP), May 2020, Barcelona, Spain. ⟨hal-02465899⟩
  • Sahar Ghannay, Yannick Estève, Nathalie Camelin. Représentations de phrases dans un espace continu spécifiques à la tâche de détection d'erreurs. XXXIIe Journées d'Etudes sur la Parole (JEP 2018), Jun 2018, Aix-en-Provence, France. ⟨hal-01757774⟩
  • Edwin Simonnet, Sahar Ghannay, Nathalie Camelin, Yannick Estève. Simulation d'erreurs de reconnaissance automatique dans un cadre de compréhension de la parole. XXXIIe Journées d'Etudes sur la Parole (JEP 2018), Jun 2018, Aix-en-Provence, France. ⟨hal-01757770⟩
  • Sahar Ghannay, Yannick Estève, Nathalie Camelin. Task Specific Sentence Embeddings for ASR Error Detection. Interspeech 2018, Sep 2018, Hyderabad, India. ⟨10.21437/Interspeech.2018-2211⟩. ⟨hal-01870864⟩
  • Edwin Simonnet, Sahar Ghannay, Nathalie Camelin, Yannick Estève. Simulating ASR errors for training SLU systems. LREC 2018, May 2018, Miyazaki, Japan. ⟨hal-01715923⟩
  • Sahar Ghannay, Yannick Estève, Nathalie Camelin. Enriching confusion networks for post-processing. Statistical Language and Speech Processing 2017, Oct 2017, Le Mans, France. ⟨hal-01585768⟩
  • Edwin Simonnet, Sahar Ghannay, Nathalie Camelin, Yannick Estève, Renato de Mori. ASR error management for improving spoken language understanding. Interspeech 2017, Aug 2017, Stockholm, Sweden. ⟨hal-01526298⟩
  • Sahar Ghannay, Benoit Favre, Yannick Estève, Nathalie Camelin. Word embedding evaluation and combination. 10th edition of the Language Resources and Evaluation Conference (LREC 2016), 2016, Portorož, Slovenia. ⟨hal-01433185⟩
  • Sahar Ghannay, Yannick Estève, Nathalie Camelin, Paul Deléglise. Acoustic word embeddings for ASR error detection. Interspeech 2016, 2016, San Francisco (CA, USA), Unknown Region. ⟨hal-01433176⟩
  • Sahar Ghannay, Yannick Estève, Nathalie Camelin, Paul Deléglise. Evaluation of acoustic word embeddings. RepEval@ACL 2016: The 1st Workshop on Evaluating Vector-Space Representations for NLP, 2016, Berlin, Germany. ⟨hal-01433181⟩
  • Yannick Estève, Sahar Ghannay, Nathalie Camelin. Recent improvements on error detection for automatic speech recognition. 1st International Workshop on Multimodal Media Data Analytics (MMDA 2016), in Conjunction with the 22nd European Conference on Artificial Intelligence, 2016, The Hague The, Netherlands. ⟨hal-01433168⟩
  • Sahar Ghannay, Yannick Estève, Nathalie Camelin, Camille Dutrey, Fabian Santiago, et al.. Utilisation des représentations continues des mots et des paramètres prosodiques pour la détection d’erreurs dans les transcriptions automatiques de la parole. 31ème Journées d’Études sur la Parole, 2016, Paris, France. ⟨hal-01450277⟩
  • Sahar Ghannay, Nathalie Camelin, Yannick Estève. Which ASR errors are hard to detect?. Workshop Errors by Humans and Machines in multimedia, multimodal and multilingual data processing (ERRARE 2015), 2015, Sinaia, Romania. ⟨hal-01433201⟩
  • Sahar Ghannay, Yannick Estève, Nathalie Camelin. Word embeddings combination and neural networks for robustness in ASR error detection. 2015 European Signal Processing Conference (EUSIPCO 2015), 2015, Nice, France. ⟨hal-01433210⟩
  • Sahar Ghannay, Yannick Estève, Nathalie Camelin, Camille Dutrey, Fabian Santiago, et al.. Combining continous word representation and prosodic features for ASR error prediction. 3rd International Conference on Statistical Language and Speech Processing (SLSP 2015), 2015, Budapest, Hungary. ⟨hal-01433203⟩
  • Sahar Ghannay, Loïc Barrault. Using Hypothesis Selection Based Features for Confusion Network MT System Combination. Third Workshop on Hybrid Approaches to Translation (HyTra), EACL 2014, 2014, Gothenburg, Sweden. ⟨hal-01433229⟩