Dr. Sangramsing N Kayte

I'm

About

I'm a Data Scientist in NLP and Speech Processing. My future vision is to build different models that involve the use of Generative Adversarial Networks, Recurrent Neural Network, Long short-term memory, etc. My research interests are broadly in the areas of NLP and Speech Processing and Machine Learning & Deep Learning as applied to NLP and Speech.

I was a European Union Postdoctoral Fellow at the University of Southern Denmark at Odense, Denmark. I obtained my Ph.D. from the Department of Computer Science and Information Technology at Dr. Babasaheb Ambedkar Marathwada University, Aurangabad, Maharashtra, India in 2017. Also in my academics, I have completed M.Phil. in Computer Science, M.Sc. in Information Technology, and Bachelor of Science degree from the same university. I have played a crucial role as project staff for the SERB-DST sponsored project in my research experience.

I specialize in handling real time and unstructured dataset, NLP modules like: Sentence Segmentation, Word Segmentation, Text Normalization, Part-of-Speech tagging, and Speech modules for Pronunciation, Prosody Prediction and Waveform generation. I have a strong interest in Machine Learning and Deep Learning, which I frequently apply to a range of NLP, TTS & ASR, Problem's and Experience in building Natural Language Processing, Text-to-Speech & Automatic Speech Recognition various languages.

Currently, I am exploring more about Deep Learning, and Generative Adversarial Networks with a focus on developing unsupervised deep learning to model human speech recognition and perception. Apart from research experience, I am also an active member of the International Speech Communication Association (ISCA) and Reviewer of IEEE Access: The Multidisciplinary Open Access Journal and Asian Journal of Research in Computer Science SCIENCEDOMAIN, Information Engineering and Applied Computing (Editorial Team Member). In my free time, I enjoy spending time with my friends, photographing architecture, exploring urban areas, enjoying a good mystery book or action movies, and doing puzzles. I am fluent in different languages like English, Hindi, Marathi, Rajasthani, Gujarati and currently learning Danish & German.

Awards

An Award, also called a distinction which is given to a recipient as a token of recognition of excellence in a particular area.

The SERB National Post-Doctoral fellowship, Govt. of India.

Got awarded grant from International Speech Communication Association ISCA

The Award of Microsoft AI Challenge, CodaLab

IEEE Honours of the Professional Development

Thesis

A long piece of writing on a specific topic, like machine learning and deep learning, or particularly one for higher college and university degrees:

Master of Science Online Toll-Plaza Management System

Machine Learning Framework Skills

Machine Learning Framework with combination of EDA i.n Exploratory Data Analysis, Planning and Feature Engineering

TensorFlow, PyTorch 100%
Keras, Theano95%
Pandas, Numpy, Matplotlib 95%
HTK, Festival, Merlin, Kaldi 90%
API90%
AWS, Azure, Google Cloud Platform, Docker, Git 95%

Machine Learning Programming Skills

Knowledge of programming languages with data structures, algorithms and OOPs concept are well diversified to learn machine learning skills.

Python (Jupyter Notebook, Spyder) 100%
R 90%
Shell Scripting85%
JavaScript & JAVA 80%
MATLAB90%
C & C++ 80%

Experience

8+ years of experience in Artificial Intelligence domain with emphasis on building solutions for Machine Learning (Supervised and Unsupervised algorithms) for Natural Language Processing, Speech Processing, Speech Synthesis, Speech Recognition, Data Science and Computer Vision area.

Data Scientist
2020 - Present

Cyrix, Copenhagen, Denmark

The approach followed to perform audio sentiment analysis on audio recordings using speech recognition and speaker recognition. The sentiment analysis for customer call data. As I have experience in speech and signal processing, with speech parameters and noise separation and speech enhancement technique, the results are quite appreciated.

Data Scientist In NLP
2019 - 2020

University of Southern Denmark

Description: The aim of the work involves the collection of Tender based database and proposes a new method of automatized text generation and subsequent classification of the European Union Tender Electronic Daily, text documents into predefined technological categories of the dataset. I implemented a neural machine learning model based on LSTM and RNN nodes for text generation and subsequent code classification.

Principle Applied Scientist-NLP (Team Lead)
2017 - 2018

Elevare Systems. AI, Hyderabad

Description: Team Lead (Natural Language Processing and Automatic Speech Recognition):- Coach is a term synonymous with mentor, guide, buddy and so many other names, demonstrating the bond between coach and an individual. Coachworks towards, overall improvement of an individual, especially, lifestyle. Just like a real-world coach, we are developing a voice based virtual assistant.

Machine Learning Research Assistant
2014 - 2017

Ministry of Science and Technology (DST), Government of India

I have worked on "Design and development of Audio Speech synthesis systems for the Indian language i.e. Hindi, Marathi, and English. The main objective of this research is to improve the accuracy of synthesized speech for these languages. The work initially started with the traditional method like- Hidden Markov Model, but later on, improved the accuracy using Deep Neural Network.

Machine Learning Research Assistant
2010 - 2013

Ministry of Science and Technology (DST), Government of India

I have worked as on Research project “Development of Multi-Resolution Analysis Technique for Early Detection of Non-Proliferative Diabetic Retinopathy without using Angiography”. We present a method for automatic detection and identify Normal, Non- Proliferative diabetic retinopathy and Proliferative retinopathy from color fundus images. The results showed a sensitivity of 95% for the classifier and specificity of 95.6%.

Academic Qualifications

In my academic qualifications, I have completed Ph.D. (Machine Learning), MPhil, MSc in Computer Science.

Postdoc in NLP
2019 - 2020

University of Southern Denmark, Odense, Denmark

I have completed a 18 month post-doc , I have worked involves the collection of Tender based database and proposes a new method of automatized text generation and subsequent classification of the European Union Tender Electronic Daily, text documents into predefined technological categories of the dataset. I have implemented a neural machine learning model based on RNN and LSTM, nodes for text generation and SVM for subsequent code classification.

Master of Philosophy (M.Phil)
2010 - 2013

Dr. Babasaheb Ambedkar Marathwada University, Maharashtra, India.

I have completed a M.Phil in Computer Science, The development of an automatic system, for the purpose of detecting anatomical and pathological features in color retinal images, with its application to diagnosis of diabetic-related eye diseases.

Thesis Available

Bachelor of Science (B.Sc)
2005 - 2008

Dr. Babasaheb Ambedkar Marathwada University, Maharashtra, India.

I have completed a Bachelor of Science (B.Sc) degree from Dr. Babasaheb Ambedkar Marathwada University, India. Bank Management System, It is used to Keep the records of clients,employee etc in Bank. The system provides the access to the customer to create an account, deposit/withdraw the cash from his account, also to view reports of all accounts present.

Doctor of Philosophy (Ph.D)
2014 - 2017

Dr. Babasaheb Ambedkar Marathwada University, Maharashtra, India

I have completed a Ph.D in Computer Science. A speech synthesis system is a computer-based system that should be able to read any text aloud with a particular language or multiple languages.

Thesis Available

Master of Science (M.Sc.)
2008 - 2010

Dr. Babasaheb Ambedkar Marathwada University, Maharashtra, India.

I have completed a M.Sc. in Information Technology, Automatic Toll Tax systems have really helped a lot in reducing the heavy congestion caused in the metropolitan cities of today. It is one of the easiest methods used to organize the heavy flow of traffic. When the car moves through the toll gate on any road, it is indicated on the RFID reader that it has crossed the clearing.

Active Professional Memberships

Membership for Professional Associations & Organizations serve the interests and knowledge sharing in a given industry or occupation.

International Speech Communication Association (ISCA) (Membership no 16479)

Institute of Electrical and Electronics Enginners (IEEE) (Membership no 94954265)

International Association of Engineers and Computer Scientists, Hong Kong (membership number 133848)

Institute of Research Engineers and Doctors (Membership no AM101000657)

Publication

To publish is to make content available to the general public and authorised. It is usually applied to text, images, or other audio-visual content, including paper.

  • Journals
  • Books
  • Book Chapter
  • International Conference
  • National Conference
  • submitted Research Artical
  1. Sangramsing N. Kayte, Monica Mundada "A Corpus-Driven Marathi Text-To-Speech System Based On The Concatenative Synthesis Approach" International Journal of Engineering Research and General Science Vol-4, Issue 1, ISSN 2091-2730, Feb-2016.   pdf

  2. Sangramsing N. Kayte, Monica Mundada "Analysis of Speech and its Fluency Disorders" International Journal of Applied Information Systems, ISSN: 2249-0868 Foundation of Computer Science, Vol-10, New York, USA, Jan-2016.    pdf

  3. Sangramsing N. Kayte, "Extraction of Speech Parameters from Speech Database using Festival" International Journal of Computer Applications (0975 8887), Vol-134, No.13, Jan-2016.    pdf

  4. Sangramsing N. Kayte, Bharti Gawali. "Analysis of Pitch and Duration in Speech Synthesis using PSOLA", s Communications on Applied Electronics 4, Published by Foundation of Computer Science, NY, USA,Feb-2016, pp.10-18.   pdf

  5. Sangramsing N. Kayte, Monica Mundada "Study of Marathi Phones for Synthesis of Marathi Speech from Text", International Journal of Emerging Research in Management and Technology ISSN:2278-9359, Issue-10, Oct-2015, IF:1.492.   pdf

  6. Sangramsing N. Kayte, Monica Mundada, Dr. Charansing Kayte "Di-phone-Based Concatenative Speech Synthesis Systems for Marathi Language", OSR Journal of VLSI and Signal Processing, Vol-5, Issue 5, Oct-2015, e-ISSN: 2319 4200, PP 76-81.   pdf

  7. Sangramsing N. Kayte, Monica Mundada, Dr. Charansing Kayte "Performance Calculation of Speech Synthesis Methods for Hindi language ", IOSR Journal of VLSI and Signal Processing, Vol-5, Issue 6, Ver. I, e-ISSN: 2319 4200, Nov-2015, PP-13-19.    pdf

  8. Sangramsing N. Kayte, Monica Mundada, Dr. Charansing Kayte "Implementation of Marathi Language Speech Databases for Large Dictionary", IOSR Journal of VLSI and Signal Processing Vol-5, Issue 6, Ver. I, e-ISSN: 2319 4200, Dec-2015, PP 40-45.   pdf

  9. Sangramsing N. Kayte, Monica Mundada, Dr. Charansing Kayte " Screen Readers for Linux and Windows Concatenation Methods and Unit Selection based Marathi Text to Speech System", International Journal of Computer Applications, Vol-130 No.14, Nov-2015.   pdf

  10. Sangramsing N. Kayte, Monica Mundada, Dr.Bharti Gawali "Transformation of feelings using pitch parameter for Marathi Speech", Journal of Engineering Research and Applications ISSN: 2248-9622, Vol. 5, Issue 11, Nov-2015, pp.120-124.   pdf

  11. Sangramsing N. Kayte, Monica Mundada, Dr.Bharti Gawali "Grapheme-To-Phoneme Tools for the Marathi Speech Synthesis", Journal of Engineering Research and Applications ISSN: 2248-9622, Vol. 5, Issue 11, Nov-2015, pp.86-92.   pdf

  12. Sangramsing N. Kayte, Monica Mundada, Dr.Bharti Gawali "Automatic Generation of Compound Word Lexicon for Marathi Speech Synthesis", IOSR Journal of VLSI and Signal Processing, Vol-5, Issue 6, Ver. II (2015), e-ISSN: 2319 4200, PP 25-30.    pdf

  13. Sangramsing N. Kayte, Monica Mundada, Dr.Bharti Gawali "Automatic Generation of Compound Word Lexicon for Marathi Speech Synthesis", Journal of VLSI and Signal Processing, Vol-5, Issue 6, Ver.II ( 2015), e-ISSN: 2319 4200, PP 25-30.   pdf

  14. Sangramsing N. Kayte, Monica Mundada, Dr.Bharti Gawali "Implementation of Text To Speech for Marathi Language Using Transcriptions Concept", Journal of Engineering Research and Applications ISSN: 2248-9622, Vol. 5, Issue 11, Nov 2015, pp.33-36.    pdf

  15. Sangramsing N. Kayte, Monica Mundada, Dr.Bharti Gawali "Rule-based Prosody Calculation for Marathi Text-to-Speech Synthesis", Journal of Engineering Research and Applications ISSN: 2248-9622, Vol. 5, Issue 11, Nov-2015, pp.33-36.    pdf

  16. Sangramsing N. Kayte, Dr.Bharti Gawali "The Marathi Text-To-Speech Synthesizer Based On Artificial Neural Networks", International Research Journal of Engineering and Technology (JET) e-ISSN: 2395-0056 Vol-2 Issue: 08| Nov-2015.   pdf

  17. Sangramsing N. Kayte, Dr.Bharti Gawali "The Prosody Subsystem and Pitch Pattern for Marathi Text To Speech Synthesis", International Journal Of Modern Engineering Research, ISSN: 22496645, Vol. 5, Iss. 12, Dec-2015.    pdf

  18. Sangramsing N. Kayte, Bharti Gawali "A Text-To-Speech Synthesis for Marathi Language using Festival and Festvox", International Journal of Computer Applications (0975 8887), Vol-132, Dec-2015.   pdf

  19. Sangramsing N. Kayte, "Text To Speech for Marathi Language using Transcriptions Theory", International Journal of Computer Applications 131(6):39-41, Published by Foundation of Computer Science, Dec-2015, NY, USA.    pdf

  20. Sangramsing N. Kayte, "Marathi Speech Recognition System Using Hidden Markov Model Toolkit", International Journal Of Modern Engineering Research, ISSN:2249 6645, Vol. 5, Iss. 12, Dec-2015.   pdf

  21. Sangramsing N. Kayte, Dr. Charansing N. Kayte, Dr.Bharti Gawali "Approach of Syllable Based Unit Selection Text-To-Speech Synthesis System for Marathi Using Three Level Fall Back Technique", Journal of Signal Processing, Vol-5, Issue 6, e-ISSN: 319 4200, Nov-2015, PP 31-35.   pdf

  22. Monica Mundada, Sangramsing N. Kayte and Dr. Bharti Gawali "Classification of Fluent and Dysuent Speech Using KNN Classier", International Journal of Advanced Research in Computer Science and Software Engineering Vol-4, Issue 9, Sep-2014.    pdf

  23. Monica Mundada, Dr. Bharti Gawali, Sangramsing N. Kayte "Recognition and Classification of Speech and its Related Agency Disorders", International Journal of Computer Science and Information Technologies, (IF 3.32).    pdf

  24. Sangramsing N. Kayte, Dr. Bharti Gawali "Marathi Speech Synthesis: A review", International Journal on Recent and Innovation Trends in Computing and Communication, ISSN: 2321-8169, Vol: 3, Iss: 6, (IF 5.83).   pdf

  1. Sangramsing N. kayte and Peter Schneider-Kamp “A Mixed Neural Network and Support Vector Machine Model for Tender creation in the European Union TED Database", KMIS 11th International Conference on Knowledge Management and Information Systems, Vienna, Austria 17-19, Sep-2019.

  2. Sangramsing N. kayte and Peter Schneider-Kamp"A Neural NLP Framework for an Optimized UI for Creating Tenders in the TED Database of the EU" , 9th Ninth International Conference on Ambient Computing, Applications, Services and Technologies, 26-Sept-2019, Porto, Portugal.

  3. Swapnil Waghmare, Sangramsing N. kayte and Ratnadeep Deshmukh "Design and Development of Stuttered and Autism Spectrum Disorder Speech Database for Marathi Language", The 22nd Conference of the Oriental, International Committee for the Co-ordination and Standardisation of Speech Databases and Assessment Techniques (COCOSDA), University of San Carlos, Gov. M. Cuenco Ave. Cebu City, Philippines 2019

  4. Swapnil Waghmare, Sangramsing N. kayte and Ratnadeep Deshmukh"Analysis of Fundamental Frequency, Jitter and Shimmer in Stuttered and Non-Stuttered Speech of Marathi Language", International Conference on Communication and Information Processing (ICCIP-2019) Available on: Elsevier-SSRN, PUNE, INDAI, pp.1-8.

  5. Sangramsing. N, Kayte and Monica Mundada, "Post-Processing using Speech Enhancement Techniques for Unit Selection and Hidden Markov Model-based Low Resource Language Marathi Text-to-Speech System", 6th Int. Workshop on Spoken Language Technologies for Under-resourced Languages(SLTU'18), Gurugram, India on 31-Aug-2018.  pdf

  6. Monica Mundada, Sangramsing. N, Kayte and Pradip K. Das, "Implementation of Concatenation Technique for Low Resource Text-To-Speech System-based on Marathi Talking Calculator” , 6th Int. Workshop on Spoken Language Technologies for Under-resourced Languages(SLTU'18), Gurugram, India on 31-Aug-2018.  pdf
  1. Sangramsing. N, Kayte "Tracking the website real-time activity using google analytics", 3rd National Conference on Emerging and Innovation Trends in Computer Science, NCEITCS-01-02 April 2014.

  2. Monica Mundada, Sangramsing N. Kayte "Classification of speech and its related agency disorders Using KNN", ISSN2231-0096 Vol-4 Number-3 Sept 2014.

  3. Sangramsing N. Kayte Bharatratna P.Gaikwad, "Design and Development of Non- Proliferative Diabetic Retinopathy Detection Technique using Image Features Extraction Techniques National", Conference in Advances in computing (NCAC13), 05-06 March 2013.
  1. Sangramsing N. Kayte and Bharti Gawali “Text-To-Speech Synthesis System Using Concatenation Technique", Scholars’ Press, eBook ISBN: 9786138831426, April 2019, pp-1-105.

  2. Sangramsing Kayte, Jaypalsing Kayte, "Non-Proliferative Diabetic Retinopathy Detection By Digital R-Images", Scholars’ Press, ISBN-13:6138830407, 2019, pp-1-108.
  1. Sangramsing N. kayte and Peter Schneider-Kamp, “A Mixed Neural Network and Support Vector Machine Model for Tender creation in the European Union TED Database", Communications in Computer and Information Science (CCIS) Series published by Springer 2019

  2. Sangramsing N. Kayte, Monica Mundada, Santosh Gaikwad and Bharti Gawali "Performance Evaluation of Speech Synthesis Techniques for English Language", Springer Science+Business Media Singapore 2016 S.C., Proceedings of the International Congress on Information and Communication Technology, Advances in Intelligent Systems and Computing 439, DOI 10.1007/978-981-10-0755-2-27.  pdf

  3. Sangramsing N. Kayte Siddharth B. Dabhade, Bharatratna P. Gaikwad, "Design and Development for Detection of Blood Vessels Microneurysms and Exudates from the Retina", Proceedings of the National Conference on Advancements in the Era of Multi-Disciplinary Systems, Elsevier Publications-2013, ISBN: 978-93-5107- 057-3, PP-394.  
  1. Sangramsing N. kayte, Monica Mundada "Low Resource Languages for Text-to-Speech Synthesis System: Survey", submitted for possible publications in ICASSP, Toronto, Canada, 2021.

  2. Sangramsing N. kayte, Monica Mundada, and Pradip K. Das "Comparison of Signal Alignment for Text-to-Speech Synthesized Speech using Dynamic Time Warping" submitted for possible publications in ICASSP, Toronto, Canada, 2021.

  3. Sangramsing N. kayte, Monica Mundada, Text-to-speech Synthesis using Generative Adversarial Network for Low Resource Indian Language" submitted for possible publications in ICASSP, Toronto, Canada, 2021.

  4. Sangramsing N. kayte, Monica Mundada, Hidden Markov Model and Deep Neural Network-based on Text-to-Speech Synthesis for Low Resource Language" submitted for possible publications in ICASSP, Toronto, Canada, 2021.

  5. Sangramsing N. kayte, Monica Mundada, A Framework for Multilingual Text-to-Speech Synthesis System for Low Resource Languages" submitted for possible publications in ICASSP, Toronto, Canada, 2021.

Speech Synthesis Sample

Speech synthesis is the task of generating speech from other modality. Rate the score as a metric with implementation of GAN and DNN

Deep Neural Networks based Synthesized samples

1. प्रयाग विश्वविद्यालय में अध्यापन के दौरान हिंदी साहित्य कोश के सम्पादन में सहयोग दिया

2. Speech Vision Laboratory

3. Language Translation Research Center

Generative Adversarial Networks based Synthesized samples

1. कहि न जाइ अति दुर्ग बिसेषी

2. देहु भगति रघुपति अति पावनि

3. अतः यहां के ग्रामीण मैथिली को अपने मातृभाषा के रूप में प्रयोग करते हैं

4.उत्तर और पश्चिम के बीच की दिशा

5. अपोलो अभियान इस यात्रा का पहला कदम

Deep Neural Networks based Synthesized samples.


1. McCoy found a stifling, poisonous atmosphere in the pent cabin.

2. This tacit promise of continued acquaintance gave Saxon a little joy-thrill.

3. He considered the victory already his and stepped forward to the meat.

4. That Longfellow chap most likely had written countless books of poetry.

5. They are greatly delighted with anything that is bright or giveth a sound.

Blog

कुछ हिस्सा हमे किस्मत से मिला तो कुछ छीना पड़ा कुछ हिस्सा हमे किस्मत से मिला तो कुछ छीना पड़ा जो किस्मत से मिला उसमें सुकून कहॉं था ! और जो छिना पड़ा उसमें अपनों का साथ भी ना था ! बेफिक्र जीयो सपनो को पूरा करो तुमने ही सिखाया था! फिर सपने भी तुम्हारे ही हाेंगे ये भी तो नहीं बताया था! सोच के देखो कभी हसरतें नई करने की तुम्हारी भी रही हाेंगी पर ये करो वो ना करो ये भी तो "दुनिया क्या सोचेगी" ये तुमने हमे सुनाया था! पर मुझे दुनिया से कुछ नही लेना देना क्यूंकि चलना मुझे तुमने ही सिखाया था! .......!

Contact

Communication with me, especially by speaking or writing.

Location:

Berlin, Germany

Call:

+45-91962211

Loading
Your message has been sent. Thank you!