Home
I am working with Calligo Technologies as Research Associate.
I have worked as Marie S Curie Early Stage Researcher in WDAqua - Web based Question Answering at Fraunhofer IAIS and Univ. of Bonn, Germany before joining SOFTRE, Tumkur. The main focus of my research work is to explore the use of semantic information in Knowledge Graphs to improve performance of Automatic Speech Recognizer.
I have completed Bachelor's in Electronics and Instrumentation (2009), and Master's in Signal Processing (2012) from Siddaganga Institute of Technology, Tumkur, India. For my master's thesis, I have worked on Distant Speech Recognition. I was associated with Medical Intelligence and Language Engineering Laboratory, (2012-2014) Indian Institute of Science, Bangalore, India to work on Text-to-Speech Synthesis project after masters. I then joined GEOSTAT team, (2014) INRIA, France, as a research intern to work on Pathological Voice Classification before joining Fraunhofer IAIS (2015), Germany.
Medium Publications
Simple, Multi-Layer, Recurrent, LSTM and Convolutional Neural Network with Tensorflow
Basic Machine Learning Algorithms For Classification With Code
Learning Java, Python, and JavaScript basics side-by-side
Publications
Ashwini Jaya Kumar, Soren Auer, Christoph Schmidt, Joachim Kohler, "Towards a Knowledge Graph Based Speech Interface," arXiv, 2017. [paper] (GLU'17, Interspeech Satellite Event)
Ashwini Jaya Kumar, Camilo Morales, Maria-Esther Vidal, Christoph Schmidt, Soren Auer, "Use of Knowledge Graph in Rescoring the N-Best List in Automatic Speech Recognition", [arXiv] 2017.
Ashwini Jaya Kumar, Christoph Schmidt, Joachim Kohler, "A knowledge Graph Based Speech Interface For Question Answering Systems", Speech Communication, Elsevier, 2017. [paper] [pdf]
Saeedeh Shekarpour, Denis Lukovnikov, Ashwini Jaya Kumar, Kemele Endris, Kuldeep Singh, Harsh Thakkar, Christoph Lange, " Question Answering on Linked Data: Challenges and Future Directions," WWW '16 Companion Proceedings of the 25th International Conference Companion on World Wide Web, 2016. [paper]
Ashwini Jaya Kumar and Khalid Daoudi, “Discrimination Between Pathological Voice Categories Using Matching Pursuit,” IEEE 4th International work conference on Bioinspired Intelligence (IOWBI' 15) June 10-12, 2015, Donostia, Spain. [pdf] [IEEE Xplore] [slides]
Khalid Daoudi and Ashwini Jaya Kumar, “Pitch-Based Speech Perturbation Measures Using a Novel GCI Detection Algorithm: Application to Pathological Voice Classification,” Interspeech, Sep 2015, Dresden, Germany. [paper][pdf]
Shiva Kumar H R, Ashwini Jaya Kumar, Rajaram B S R, Ramakrishnan A G, “MILE TTS for Tamil and Kannada” Blizzard Challenge 2013.[pdf] http://www.festvox.org/blizzard/blizzard2013.html
Ashwini Jaya Kumar, Ramaswamy Kumaraswamy, “Speech Enhancement Techniques for Distant Speech Recognition”, Journal of Intelligent Systems. May 2013. [pdf]
Ashwini Jaya Kumar, Ramaswamy Kumaraswamy, “Suppressing Unwanted Reflections Smeared in the Distant Speech Signal Using Long-Term Linear Prediction,” Proc. of International conference on Communication, VLSI and Signal Processing -Feb 20-22, 2013, pp 171-176, Tumkur, India. [pdf]
Presentations
"Speech Recognition and Understanding ", March 2017, WDAqua workshop, Fraunhofer IAIS, Germany. [slides]
"Using novel F0 estimation algorithm for pathological voice classification" and "Discrimination between pathological voice categories using Matching Pursuit", July 2015, MILE lab, Indian Institute of Science, Bangalore, India. [slides]
"Brief Introduction to Distant Speech Recognition and Speech Synthesis", March 2014, Geostat team, INRIA, Bordeaux, France. [slides]
Summer School attended
6th ESWC Summer School, 5th-10th Sept 2016, Dubrovnik, Croatia. [link]
Web Intelligence Summer School 31st Aug-4th Sep 2015 Saint-Etienne, France. [link]
Statistical Parametric Speech Synthesis, 22nd-25th Feb 2013, WISSAP, IIT Madras, Chennai, India.
Other activities
Sub Reviewer - 12th ITG Conference on Speech Communication 2016
Conference Organisation - CSCUBS'16
Bachelors/Masters thesis Co-supervision
Disclaimer
Copyright and all rights of the documents made available here is reserved. These works may not be reposted without the permission of the copyright holder.