Zeeshan Ahmed
Biography
Zeeshan Ahmed is a PhD student in MUSTER research group in UCD school of computer science and informatics. He received his BS in Computer Science from Department of Computer Science,University of Karachi, Pakistan. In 2007, he was awarded with European Union Erasmus Mundus scholarship for studying in Language and Technology program. He completed his first year of Master in Charles University of Prague,Czech Republic. During his first year, his studies focused on statistical methods in Speech and Language Processing. For 2nd year master, he attended University of Nancy,France. Where his focus was on Knowledge based approaches to language processing. In 09/10, he was awarded with SFI/CNGL scholarship to further enhance his skills in Speech Processing.
Research
The focus of my research is the semi-integrated approach to speech translation for large vocabulary and open domains applications. When integrating ASR and MT components in a semi-integrated approach, the biggest question raised is what type of information needs to be passed to MT from ASR? Previously, n-best word list and word lattice has been tried but in this research, I am interested in looking at what other form of information can be extracted from ASR and pass onto MT for ef ective translation. The other aspect of my research is to look at the speed at which translation is performed because increasing the size of the input has direct impact on the translation system performance. Furthermore, MT and TTS integration would also be considered for faster performance.
Furthermore, a speech translation system requires a di erent type of data resource for its development e.g. a source speech corpus for ASR, a parallel corpus for MT and a target speech corpus for speech synthesis. Nowadays, each of these resources can be obtained easily but the problem is that for speech translation these resources need to be coherent i.e. the domain and vocabulary size of ASR, MT and TTS corpora should be same. One of the objectives of my research is to minimize this dependency.
Publications
CONFERENCE:
2012:
Eva Szekely, Zeeshan Ahmed, Joao P. Cabral and Julie Carson-Berndsen. WinkTalk: a demonstration of a multimodal speech synthesis platform linking facial expressions to expressive synthetic voices. In Third Workshop on Speech and Language Processing for Assistive Technologies (SLPAT),Montreal, Quebec, Canada (2012)
Mark Kane, Zeeshan Ahmed and Julie Carson-Berndsen. Underspecification in Pronunciation Variation. In Processing of International Symposium on Automatic Detection of Errors in Pronunciation Training, Stockholm, Sweden (2012).
Joao Paulo Cabral, Mark Kane, Zeeshan Ahmed, Mohamed Abou-Zleikha, Eva Szekely, Amalia Zahra, Kalu Ogbureke, Peter Cahill, Julie Carson-Berndsen and Stephan Schlogl. Rapidly Testing the Interaction Model of a Pronunciation Training System via Wizard-of-Oz. In : The 8th International Conference on Language Resources and Evaluation (LREC),Istanbul,Turkey(2012).
2011:
Jie Jiang, Zeeshan Ahmed, Julie Carson-Berndsen, Peter Cahill and Andy Way. 2011. Phonetic Representation-Based Speech Translation. In Proceedings of Machine Translation Summit XIII, Xiamen, China
Zeeshan Ahmed, Julie Carson-Berndsen: Automatic Rule Extraction for Modeling Pronunciation Variation. In: 12th International Conference on Intelligent Text Processing and Computational Linguistics, Tokyo, Japan(2011).
Peter Cahill, Udochukwu Ogbureke, João Cabral, Eva Szekely, Mohamed Abou-Zleikha, Zeeshan Ahmed, Julie Carson-Berndsen. UCD Blizzard Challenge 2011 Entry.
2010:
Zeeshan Ahmed, Julie Carson-Berndsen: Modeling pronunciation of OOV words for speech recognition. In: Thirteenth Australasian International Conference on Speech Science and Technology, Melbourne, Australia(2010).
WORKSHOP :
2012:
Zeeshan Ahmed, João Cabral, Udochukwu Ogbureke, Julie Carson-Berndsen. HMM-Based Expressive Speech Synthesiser for Urdu Language: Workshop on Innovation and Applications in Speech Technology, March 9-10, 2012, Dublin, Ireland.
Eva Szekely, Joao Cabral, Zeeshan Ahmed, Julie Carson-Berndsen. Eyebrows speak louder than words - linking facial expressions to synthetic voice: Workshop on Innovation and Applications in Speech Technology, March 9-10, 2012, Dublin, Ireland.
Teaching
2012: Teaching Assitant:
Compiler Construction (COMP30330)
2011: Teaching Assitant:
Compiler Construction (COMP30330)
Foundation of Computing (COMP30010)
Object Oriented Programming (COMP30070)
Design Patterns with Ruby (COMP40070)
Managing Software in Production (COMP41420)
2010:
Formal Foundation (COMP10070)
C/C++ Programming I (COMP10110, COMP30400)
Algorithm Problem Solving (COMP10030)
- Login to post comments