Nia Peters

Research Interests

My current research endeavors include exploring speech and natural language processing and modeling strategies within multi-human, multi-machine teaming interactions. This research focuses on the algorithms needed for machines to communicate efficiently and effectively within these interactions, as well as the implications these communication strategies have on human teammates.

keywords:  spoken language processing, human-machine interaction, natural language processing, artificial intelligence, human factors engineering

Project(s)

Context-Aware Communication Interfaces (2017-2020)

Human-machine teaming aims to meld human cognitive strengths and the unique capabilities of smart machines to create intelligent teams. One major problem with these teams is a lack of communication skills on the part of the machine such as the inability to know when and how to communicate information to human teammates using spoken language. One scientific gap is our ability to measure and evaluate task interruptibility or “when” a machine should communicate information to human teammates engaged in an ongoing task.  With an agent that must monitor and attend to all speech communication streams to encode appropriate times to engage with teammates communicating and working on an ongoing task, formalized models could be used within the literature.

The Air Force Research Laboratory Battlespace Acoustic Branch is exploring this issue through an intelligent interruption system integration into simple and complex human-machine collaborative teaming interactions via the Collaborative Communication Interruption Management System (C-CIMS) project.  The objective of C-CIMS is to leverage low-level spoken language from collaborative human teaming interactions and to infer candidate interruption timings.  This project will  (1) support the development of a working dataset to simulate collaborative human-machine interactions where C-CIMS will be integrated (2) design and evaluate modeling strategies for C-CIMS to make appropriate interruption decisions, and  (3) evaluate the implications of these decisions on human performance, task resumption times, and human affect state during team interactions.

Speech Technology for Court Transcriptions (2020 – Present)

Stenographers and court reporters are responsible for capturing court proceedings and depositions into written text.  Stenograph LLC is responsible for the technology these professionals use to do their job.  With the emergence of burgeoning technologies such as speech and natural language systems like Nuance, Alexa, and Hey Google, there is an opportunity to leverage the same technologies to allow court reports and stenographers to their job more efficiently.  Currently I work on technologies that will integrate into the systems these professionals already use.  The objective is to provide “hypothesis” of what was said during court proceeding from the speech system so the reporter can correct both offline and in real-time.  I work on different components of this technology including automatic speech recognition or the conversion of speech to text, diarization or the technology that captures who spoke when, and finally grammar and post-processing or the mapping between raw text and appropriate punctuation and capitalization.

Invited Talks

28 April 2021 – Air Force Research Laboratory INSPIRE – Toward Natural Human-Machine Communication – Peters, Nia 

22 October 2020 – Air Force Research Laboratory Naval Research Laboratory Speaker – Collaborative Communication and Intelligent Interruption Management – Peters, Nia S.

6 October 2020 – Human-Autonomy Teaming Scientific Advisory Board Speaker – Context-Aware Communication within Human-Autonomy Interactions – Peters, Nia S.

20 August 2020 – AFRL SBIR/STTR HBCU Collider Panel Discussion Participant – Career Development – Peters, Nia S.

13 April 2020 – 711th Human-Performance Wing Commander Immersion Speaker – Context-Aware Human Communication – Peters, Nia S.

15. May 2020 – North Carolina A&T University College of Engineering Guest Speaker – Exploring Human Communication Processes to Inform Intuitive Human-Machine Communication – Peters, Nia S.

28 January 2020 – Air Force Research Laboratory and Netherlands Aerospace Center Technical Visit Speaker –  A Collaborative Communication Interruption Management System (C-CIMS) – Peters, Nia S. 

22 October 2019 – AFRL Airman Systems Directorate Scientific Advisory Board Review – Exploring Human Communication Processes to Inform Intuitive Human-Machine Communication – Peters, Nia S.

13 May 2019 – Air Force Office of Scientific Research Laboratory Briefing – Collaborative Communication Interruption Management System (C-CIMS) – Peters, Nia S.

19 April 2019 – Air Force Research Laboratory (AFRL) Division Leadership Meeting Speaker – Collaborative Communication Interruption Management System (C-CIMS)within Human Machine Teaming – Peters, Nia S.

20 February 2019 – AFRL 711th Human Performance Wing (HPW) Chief Scientist Immersion Briefing – Machine Communication Etiquette within Collaborative Human-Machine Teaming – Peters, Nia S.

2 May 2018 – AFRL 711th Human-Performance Wing Behavioral Science Brown Bag Speaker – A Collaborative Communication Interruption Management System (C-CIMS): Modeling Interruption Timings via Prosodic & Topic Modeling for Human-Machine Team – Peters, Nia S.

Research & Scholarship

Teaching & Mentoring