Our Team - Bangla Language Computing Research

Dr. Shammur Absar Chowdhury is currently a Post-Doctoral Researcher at QCRI. Prior to that was also a Post-Doctoral Researcher at  the University of Trento where she was investigating grammatical structure of a language, and  analyzing conversational behavior (e.g., who competes in a conversation) using deep neural network. Dr. Chowdhury received her PhD in Computer Science from the ICT Doctoral School at the University of Trento, Italy. She has received her B.Sc. in Electronics and Communication Engineering along with B.Sc. in Computer Science from the BRAC University, Bangladesh. During her PhD she has contributed to the research of spoken conversation analysis and human-machine interaction while focusing on the discourse modeling of overlapping speech, silences and predicting observed behavior (e.g., speaker satisfaction, personality, emotional coordination, and turn-taking behaviors) of the speakers from speech and its transcription (text).  She has also been part of many European Union Projects (e.g., SENSEI, Portdial). Her developed Speech Recognition System for Spoken Conversations is currently the best system for Italian. After her undergrad, she has joined as a faculty member at BRAC University, and alongside she was also a technical lead in text to speech project at CRBLP, which was funded by Prothom-alo newspaper. She has also developed the first continuous speech recognition system for Bangladeshi Bangla. With novel contributions she has published many research papers in international conferences and journals.

S. M. Murtoza Habib is a Software Engineer at Verizon Connect. He has been working on Test Driven Development, Refactoring, C#, and Angular. Over the time has gained experience working in the Agile environment both in Scrum and Kanban. He received his B.Sc. in computer science and engineering from BRAC University, Bangladesh. As a researcher at BRAC University, he was mainly worked on Corpus Development, Microsoft localizaiton project, and Speech Synthesis (Text to Speech) from 2007 to 2010, a first contribution for Bangladeshi Bangla. It was a part of Pan Localization project (www.panl10n.net).

Dr. Md. Abul Hasnat is a Senior Data Scientist at VINCI (Cyclope Team), Paris, France. Currently, he is working on face recognition using the deep convolutional neural network (CNN) architecture. Previously, he was a postdoctoral researcher at the Data Mining & Decision team in the ERIC Lab, University Lumière Lyon 2, France. He worked on the ANR-ImagiWeb project to develop novel evolutionary clustering method for analyzing and understanding the temporal dynamics of social media data. He received his BSc degree from BRAC University, Bangladesh, in 2006, the MSc degree from Erasmus Mundus Color in Informatics and Media Technology (CIMET) in 2011, and the PhD degree in “image, vision and signal ” from the Laboratoire Hubert Curien UMR CNRS 5516, Université Jean Monnet Saint-Etienne, France, in 2014. In his PhD thesis, he developed novel unsupervised methods for analyzing 3D and RGB-D (joint color+depth) data captured by the Microsoft Kinect sensor. He started his professional career as a researcher at the Center for Research on Bangla Language Processing (CRBLP), BRAC University, where he conducted research on different projects and developed the first complete Bangla OCR software. His research activities are mainly focused in the area of computer vision, image processing, pattern classification, machine learning, and data mining. He has over 20 research papers, published in international conferences and journals.

Dr. Firoj Alam is Post-Doctoral Researcher at the Qatar Computing Research Institute (QCRI), Qatar. At QCRI he is contributing to the AIDR (Artificial Intelligence for Digital Response, http://aidr.qcri.org) project, specifically investigating semi-supervised deep learning algorithms to categorize social media messages related to emergencies, disasters, and humanitarian crises. His research interests include Applied Machine Learning, Deep Learning, Natural Language Processing, Affective Behavior and Speech Processing. Firoj received his PhD in Computer Science from the ICT Doctoral School of the University of Trento, Italy, and M.Sc in Human Language Technology from the same University. He received his B.Sc. in computer science and engineering from BRAC University, Bangladesh. As a researcher at BRAC University, he mainly worked on Speech Synthesis (Text to Speech) from 2007 to 2010, a first contribution for Bangladeshi Bangla. It was a part of Pan Localization project (www.panl10n.net). Other Research work at CRBLP include developing lexicons, speech corpus and core NLP resources. During his PhD, he worked on Behavioral Signal Processing (Empathy, Emotion, Personality traits) from Speech and Text. He has been part of EU conversation summarization project – SENSEI (www.sensei-conversation.eu). During his master’s internship at FBK, Italy, in 2011, he worked on information extraction, particularly, Named Entity Recognition, which was a part of live memories project  (www.livememories.org). He has over 20 research papers, published in international conferences and journals.

Rabia Sultana Ummi is a Consultant at Cognitive Insight Limited. Her research interest includes Bangla Computing Research, specifically, Bangla POS tagging, stemmer, lemmatizer, and corpus development. During her time at CRBLP, she contributed to Bangla Text Speech (kotha and shuti), development of Bangla pronunciation dictionary. She finished her Master of Science in Computer Science from Independent University and Bachelor of Science in Computer Science from BRAC University. Her thesis work shed light on Bangla Stemmer and Bangla Machine translation research.

Muttakinur Rahman Chowdhury is a Senior Software Engineer at Cognitive Insight Limited. He has been working with Building AI technologies focusing on Bangla Language, particularly, corpus development tools, annotation platforms and maintaining backend systems. He has worked with many different technologies and gained experience of how to design a software system. He has significant experience in Docker container technology, Kubernetes, Microservice architecture, Backend development, Programming, Cloud deployment Automation, DevOps, System automation, systems analysis and design, Web service development, Web application API. He has finished his Bachelor in computer science and engineering from BRAC University and has gained research and development experiences significantly while working with CRBLP team.

Md. Arid Hasan is a Research Programmer at Cognitive Insight Limited. He has been working with statistical and neural machine translation system focusing on Bangla-English Language pair, particularly, corpus cleaning, running different experiments with different parameters and measure both performances and challenges. His research interest includes Natural Language Processing and Machine Learning, especially, statistical machine translation, neural machine translation, and corpus development. He has experience in Python, OpenNMT, Pytorch, Web application development. He has completed his Bachelor in computer science and engineering from Daffodil International University and has gained research and development experiences during his undergrad Thesis and working with 'DIU NLP & ML Lab'. His undergrad thesis work mainly focused on the Collaborative Parallel Corpus development tool and Machine Translation systems.