Research Interest

Broadly, I am interested in applying machine learning algorithms to different problems in speech, text and image domains.

My Ph.D. research is on unsupervised audio segmentation based on specific events. In this, I am working towards improving the speaker boundaries in conversational speech for the speaker diarization systems. In my work, I am trying to incorporate speaker discriminative representation learning within an unsupervised clustering framework. These representations can be obtained from the neural networks or other discriminative classifiers. I am also working towards continual incremental learning in speaker diarization and Knowledge Distillation for E2E ASR model compression.

Publications

  • Google Scholar (Updated list on Google scholar)         DBLP


    • Journal Paper:

  • • Nauman Dawalatabad, Srikanth Madikeri, C. Chandra Sekhar, Hema A. Murthy,
    "Novel Architectures for Unsupervised Information Bottleneck based Speaker Diarization of Meetings,"
    in IEEE/ACM Transactions on Audio, Speech, and Language Processing, 29, 14-29, 2021. [paper] [slides] [video]


    • Conference Paper:

  • • Nauman Dawalatabad, Mirco Ravanelli, François Grondin, Jenthe Thienpondt, Brecht Desplanques, Hwidon Na
    "ECAPA-TDNN Embeddings for Speaker Diarization,"
    in Proc. of INTERSPEECH, ISCA, Brno, Czech Republic, 2021. [paper] [code] [model]

  • • Nauman Dawalatabad, Jilt Sebastian, Jom Kuriakose, C Chandra Sekhar, Shrikanth Narayanan, Hema A. Murthy,
    "Front-end Diarization for Percussion Separation in Taniavartanam of Carnatic Music Concerts,"
    ArXiv preprint arXiv:2103.03215, 2021. [paper]

  • • Nauman Dawalatabad, Srikanth Madikeri, C. Chandra Sekhar, Hema A. Murthy,
    "Incremental Transfer Learning in Two-pass Information Bottleneck based Speaker Diarization System for Meetings,"
    in IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Brighton, UK, May, 2019. [paper]

  • • Nauman Dawalatabad, Jom Kuriakose, C. Chandra Sekhar, Hema A. Murthy,
    "Information Bottleneck based Percussion Instrument Diarization System for Taniavartanam of Carnatic Music Concerts,"
    in Proc. of INTERSPEECH, ISCA, Hyderabad, India, September, 2018. [paper]

  • • Sakshi Verma, K L Prateek, Karthik Pandia, Nauman Dawalatabad, Rogier Landman, Jitendra Sharma, Mriganka Sur, Hema A. Murthy,
    "Discovering Language in Marmoset Vocalization,"
    in Proc. of INTERSPEECH, ISCA, Stokholm, Sweden, August, 2017. [paper]

  • • Nauman Dawalatabad, Srikanth Madikeri, C. Chandra Sekhar and Hema A. Murthy,
    "Two-Pass IB based Speaker Diarization System using Meeting-Specific ANN based Features,"
    in Proc. of INTERSPEECH, ISCA, San Francisco, USA, September, 2016. [paper]

Professional Experience

Project Associate

(Jan 2014 - Feb 2016)

I was involved in the project on Speaker Recognition and Speaker Diarization sponsored by Defence Research & Development Organisation (DRDO), India. I was responsible for parallelizing UBM computations (on Nvidia GPUs using CUDA-C) and developing the speaker diarization system.



Teaching Assistant

(Jan 2016 - Present)

I have been teaching assistant for the following courses.

• Pattern Recognition and Machine Learning (Fall 2019, 2018, 2017)
• Deep Learning (Spring 2019)
• Kernel Methods for Pattern Analysis (Spring 2017)
• Foundations of Computer Sytem and Design (Fall 2017)
• Speech Technology (Spring 2016)

Mentorship

(2016 - 2019)

Throughout my Ph.D. journey I was fortunate to mentor some of the brilliant students from IIT Madras and other Indian institutes with their projects/research at IIT Madras.

• Hans Tiwari - Intern 2019
  Topic: "Deep Learning based Image Recognition"
• Bhargav Dindukurthi - Intern 2018
  Topic: "HMM/GMM based speaker Diarization"
• Yaswanth Sai - Master's student at IIT-M, 2017-2018
  Topic: "End-to-End Automatic Speech Recognition"
• Sarath Chandra - Master's student at IIT-M, 2017-2018
  Topic: "Deep Embeded Clustering"
• Prateek Kotha - Master's student at IIT-M, 2016-2017
  Topic: "Speaker Diarization"
• Sakshi Verma - Master's student at IIT-M, 2016-2017
  Topic: "Diarization of Marmoset Vocalization"
• Sakil Ansari - Intern 2016
  Topic: "Handwriting Recognition"
• Jessie Sravya - Intern 2016
  Topic: "Continuous Speech Recognition"

About Myself

I am a Ph.D. candidate in the Department of Computer Science and Engineering at Indian Institute of Technology Madras, India. I joined IIT Madras in 2014 for Master of Science (by research) programme with Prof. C Chandra Sekhar where I was working on Speaker Verification and Speaker Diarization. I continued my Ph.D. from 2016 jointly with Prof. C Chandra Sekhar and Prof. Hema A Murthy on the topic of Speaker Diarization. I also collaborate with Dr. Srikanth Madikeri, Idiap Research Institute, Switzerland.

For my research work, I have been fortunate to collaborate with Prof. Shrikanth Narayanan (USC, USA) and Prof. Mriganka Sur (MIT, USA). During my Ph.D., I have mentored many students from IIT M and other Indian institutes with their projects/research works. I was also a visiting research student at Mila - Quebec AI Institute, Canada, under the supervision of Prof. Yoshua Bengio and Dr. Mirco Ravanelli.



Let us be social

       
               

Nauman Dawalatabad

Research Scholar, IIT Madras

nauman@cse.iitm.ac.in

Awards and Achievements

• Recipient of STAR Teaching Assistant (STAR-TA) award, Department of Computer Science and Engineering, IIT Madras, 2020.

• Recipient of Kris Gopalakrishnan Endowment Student Travel grant award to visit ICASSP-2019, Brighton, UK, 2019.

• Recipient of STAR Teaching Assistant (STAR-TA) award, Department of Computer Science and Engineering, IIT Madras, May, 2018.

• 1st prize winner in 5-Minute PhD Thesis (5MPT) competition supported by International Speech Communication Association (ISCA) at Summer School on Speech Signal Processing (S4P), DA-IICT, India, July, 2017.

• Recipient of the International Speech Communication Association (ISCA) travel grant award 2016 to visit Interspeech at San Francisco, USA, 2016.

• Secured All India Rank 21 in part A of Joint Entrance Screening Test (JEST) for Computer Science conducted by the Institute of Mathematical Sciences (IMSc), India, 2013.

• Secured All India Rank 6 in SRMEE (Computer Science), an entrance exam for post-graduation conducted by SRM University, India, 2013.