Profile picture

Anupama Chingacham

PhD Candidate(SFB 1102)

Saarland University(UdS)
Spoken Language Systems (LSV)
Chair of Computational Linguistics (CoLi)


achingacham [at] lsv [dot] uni-saarland [dot] de

Building C 7.1
Saarland University
Saarbrücken
Germany, 66123

About

Hello World! I am Anupama Chingacham, a doctoral student in the Department of Computer Science at Saarland University, Germany. Under the supervision of Prof. Dr. Dietrich Klakow and Prof. Dr. Vera Demberg, my current research is on paraphrasing to improve speech perception in noise . Recently, we proposed a novel framework for noise-adaptive spoken dialogue systems, which utilizes paraphrases to generate better intelligible utterances. My research interest, however, spans multiple areas in computational linguistics, especially in the landscape of employing large-scale pre-trained language models (PLMs) for low-resource settings and controlled text generation problems.

News

  • June 2023 Invited talk at EXPRESSION team in IRISA, France.
  • Jan 2023 Invited talk at COG-MHEAR team in Edinburgh Napier University, United Kingdom.
  • Jan 2023 Poster presentation at IEEE SLT, Qatar.
  • Sept 2021 Our work received a Best Student Paper award at Interspeech!
  • Sept 2021 Paper presentation at Interspeech, Czechia (hybrid).
  • Dec 2018 Paper presentation at CLiC-it, Italy.

Publications

  
A Data-Driven Investigation of Noise-Adaptive Utterance Generation with Linguistic Modification
Chingacham, Anupama and Demberg, Vera and Klakow, Dietrich
Ninth IEEE Spoken Language Technology Workshop, 2023


Exploring the Potential of Lexical Paraphrases for Mitigating Noise-Induced Comprehension Errors
Chingacham, Anupama and Demberg, Vera and Klakow, Dietrich
22nd INTERSPEECH Conference, 2021


Generalizing Representations of Lexical Semantic Relations
Chingacham, Anupama and Paperno, Denis
Fifth Italian Conference on Computational Linguistics, 2018

Projects

  • PiN: Paraphrases in Noise Dataset - annotated with sentence-level intelligibility of sentential-paraphrases in noisy environments.
  • SiN: Synonyms in Noise Dataset - annotated with word intelligibility of synonyms in noisy listening setups.
  • Relational Representations Python, PyTorch - an extended Skip-gram model to represent semantic relation between two lexical units.

A growing list of my volunteerism

  • Feb 2019 - June 2022 PhD Representative @ SFB 1102

Experience

  • Doctoral studies

    University of Saarland, Germany

    Oct 2018 - present

    Pi-SPIN: Paraphrase to improve Speech Prception in Noise.

  • Research Intern

    Loria, CNRS, France

    Feb 2018 - July 2018

    Explored an unsupervised method to represent lexical relations.

    Published our findings at CLiC-it conference.

  • MSc Natural Language Procesing

    Lorraine University, France

    Sept 2017 - July 2018

    Acquired knowledge on data modeling techniques and its application on NLP problems like for semantic parsing.

  • Networking Engineer Consultant

    Cisco Systems Pvt Ltd (India)

    Sept 2014 - Aug 2017

    Automated applications migration plan generation.
    Orchetrated application migration for multiple service providers.

  • Intership

    Cisco Systems Pvt Ltd (India)

    June 2014 - Sept 2014

    Successfully completed CCNA, CCNP certifications.

  • Masters in Software Engineering (5 Year Integrated)

    Vellore Institute of Technology, India

    2009 - 2014

    A recipient of Top 10 Academic Performers award for 4 years.
    CGPA: 9.2/10




Website built from Marcos Horro's template