| May 2023-present |
| Title |
Development of ASR systems for ATC systems |
| Supervisor |
Dr. Petr Motlicek, Idiap Research Institute, Martigny, Switzerland |
| Funding Agency |
Eurocontrol, Europe |
| Working as a Postdoctoral Researcher in a project that requires the development of automatic speech recognition, concept extraction and text-to-speech synthesis solutions for air traffic controller communication to be used as automatic training simulators to be deployed across Europe. Challenges in the task involve making the ASR module robust to rare and unknown (out-of-vocabulary) words that may arise during tech transfer across different countries in Europe. |
|
| Apr 2022-Apr 2023 |
| Title |
Development of scalable speaker recognition systems for Indian languages [Website] |
| PI |
Prof. S. R. Mahadeva Prasanna, Dept. EE, IIT Dharwad, Karnataka, India |
| Funding Agency |
MeitY, Govt. of India |
| Working as a Senior Project Engineer in a MeitY, Govt. of India sponsored project for developing speaker recognition systems for Indian languages. Entrusted with developing an open-source Python-based speaker recognition toolkit that consists of state-of-the-art methods for encouraging further research in the community. Developing the project website, system demo prototypes, and web-based portals to provide hands-on experience of the systems to beginners. |
|
| Jan 2016-Jan 2023 |
| Title |
Analysis of Speech and Music Content for Movie Genre Classification (Ph. D. Thesis) |
| Supervisors |
Prof. S. R. Mahadeva Prasanna, Dept. of EE, IIT Dharwad, Karnataka, India
Dr. Prithwijit Guha, Dept. of EEE, IIT Guwahati, Assam, India |
| Funding Agency |
Visvesvaraya Ph.D. Scheme, MeitY, Govt. of India |
| Proposed magnitude spectrum-based features for speech and music classification.
Explored phase information for speech and music classification Proposed harmonic-percussive source separation features with classifiers based on multi-task learning framework for speech overlapped with music detection Proposed a movie-trailer genre classification system using speech and music information as features.
|
|
| Jul 2013-May 2014 |
| Title |
Redundancy Reduction in Wireless Sensor Networks (M. Tech. Thesis) |
| Supervisor |
Prof. Subrata Gupta, Dept. of EE, NIT Raipur, Chhattisgarh, India |
| Funding Agency |
MHRD, Govt. of India |
| Proposed an algorithm for determining overlap between adjacent sensor sensing area. Proposed method to shut down redundant nodes in a Wireless Sensor Network for energy saving.
|
|
| Jul 2010 - May 2011 |
| Title |
N-Coins Problem: An Algorithmic Solution (B. Tech. Project) |
| Supervisor |
Prof. Rajat Kanti Pal, Dept. of IT (renamed CSE), Assam University Silchar, Assam, India |
| Implemented an algorithm for solving the N-coins problem in C. The algorithm determines a counterfeit coin among a set of N identical coins using a minimal number of comparisons.
|
|
