Shahab Jalalvand - PhD

Senior research scientist

Interactions Corp.

New Providence, NJ

Bio:

Shahab is a senior researcher in automatic speech recognition (ASR) and natural language processing (NLP) at Interactions Corp, US, where he joined in 2017. Prior to joining Interactions, he was a researcher and developer at PerVoice Inc, Italy in 2016. Prior to PerVoice, he was a PhD student researcher at FBK, Italy in 2012-2015. He holds a PhD degree in computer science from ICT Doctoral school at the University of Trento, Italy (2017). He obtained his MSc degree in artificial intelligence from Iran University of Science and Technology, Iran (2012) and BSc degree in computer engineering from the same university (2009).

His research has been focused on ASR, NLP and spoken language understanding (SLU) systems and the applications of machine learning algorithms to improve these systems. He has published over 15 journal and conference papers on these topics. He has been an active member of ISCA group from 2012.

Shahab has developed several techniques to automatically predict the quality of ASR transcriptions when no human reference is available, "TranscRater". He has developed techniques to rapidly adapt the ASR and SLU systems and improve their robustness towards speaker/accent variation and domain variation.

Currently, Shahab is conducting his research on spoken language understanding for customer service in different domains, specially when limited in-domain data is available. Moreover, he is developing techniques to improve the language models in the SLU systems for customer service applications.

Research interests:

Automatic Speech Recognition
Natural Language Processing
Spoken Language Understanding
Machine Learning
Language Modeling
Deep Neural Network

Publications

Journals and Theses

Jalalvand, Shahab; Ljolje, Andrej; Bangalore, Srinivas; Automatic Data Expansion for Customer-care Spoken Language Understanding, arXiv:1810.00670v1 [cs.CL], pp. 1-10, 2018
Jalalvand, Shahab; Negri, Matteo; Falavigna, Daniele; Matassoni, Marco; Turchi, Marco; Automatic Quality Estimation for ASR System Combination, Computer Speech & Language, pp. 214-239, 2018
Falavigna, Daniele; Matassoni, Marco; Jalalvand, Shahab; Negri, Matteo; Turchi, Marco , DNN Adaptation by Automatic Quality Estimation of ASR Hypotheses , Computer Speech & Language , pp. 585-604 , 2017
Jalalvand, Shahab , Automatic Speech Recognition Quality Estimation, PhD thesis, University of Trento, Tento, Italy , 2017

Conferences

Matassoni, Marco; Ravanelli, Mirco; Jalalvand, Shahab; Brutti, Alessio; Falavigna, Daniele , The FBK System for the CHiME-4 Challenge , 4th CHiME Speech Separation and Recognition Challenge , San Francisco, US , 2016
Jalalvand, Shahab; Negri, Matteo; Turchi, Marco; de Souza, Jose GC; Daniele, Falavigna; Qwaider, Mohammed RH , Transcrater: a Tool for Automatic Speech Recognition Quality Estimation , Proceedings of ACL-2016 System Demonstrations , pp. 43-48, Berlin, Germany , 2016
Farajian, M Amin; Chatterjee, Rajen; Conforti, Costanza; Jalalvand, Shahab; Balaraman, Vevake; Di Gangi, Mattia A; Ataman, Duygu; Turchi, Marco; Negri, Matteo; Federico, Marcello , FBK's Neural Machine Translation Systems for IWSLT 2016 , Proceedings of the ninth International Workshop on Spoken Language Translation , Seattle, WA, US , 2016
Jalalvand, Shahab; Negri, Matteo; Falavigna, Daniele; Turchi, Marco, Driving rover with segment-based asr quality estimation, Proceedings of ACL-2015, pp. 1095-1105, Beijing, China, 2015
Jalalvand, Shahab; Falavigna, Daniele , Stacked Auto-encoder for ASR Error Detection and Word Error Rate Prediction , Sixteenth Annual Conference of the International Speech Communication Association , Dresden, Germany , 2015
Jalalvand, Shahab; Falavigna, Daniele; Matassoni, Marco; Svaizer, Boosted Acoustic Model Learning and Hypotheses Rescoring on the CHiME-3 Task ,Automatic Speech Recognition and Understanding (ASRU), 2015 IEEE Workshop on , pp. 409-415, Scottsdale, Arizona, US, 2015
Jalalvand, Shahab; Falavigna, Daniele , Direct Word Graph Rescoring using A* Search and RNNLM , Fifteenth Annual Conference of the International Speech Communication Association , Singapore , 2014
BabaAli, Bagher; Serizel, Romeo; Jalalvand, Shahab; Falavigna, Daniele; Gretter, Roberto; Giuliani, Diego , Fbk@ IWSLT 2014 ASR Track , Proceedings of the International Workshop on Spoken Language Trnaslation (IWSLT) , Lake Tahoe, California, US, 2014
Jalalvand, Shahab , Improving Language Model Adaptation using Automatic Data Selection and Neural Network , Proceedings of the Student Research Workshop associated with RANLP 2013 , pp. 86-92, Hissar, Bulgaria, 2013
Jalalvand, Shahab; Falavigna, Daniele , Parameter Optimization for Iterative Confusion Network Decoding in Weather-Domain Speech Recognition , IWSLT 2013 , Heidelberg, Germany , 2013
Ivanov, Alexei V; Jalalvand, Shahab; Gretter, Roberto; Falavigna, Daniele , Phonetic and Anthropometric Conditioning of MSA-KST Cognitive Impairment Characterization System , Automatic Speech Recognition and Understanding (ASRU), 2013 IEEE Workshop on , pp. 228-233, Olomouc, Czech Republic, 2013
Jalalvand, Shahab; Akbari, Ahmad; Nasersharif, Babak , A Classifier Combination Approach for Farsi Accents Recognition , 20th Iranian Conference on Electrical Engineering (ICEE), pp. 716-720, Tehran, Iran, 2012
Jalalvand, Azarakhsh; Minaei, Behrouz; Atabaki, Golnaz; Jalalvand, Shahab , A New Interestingness Measure for Associative Rules based on the Geometric Context , 3rd International Conference on Convergence and Hybrid Information Technology (ICCIT) , pp. 199-203 , 2008