Research

We do industrial, academic and open-source research on diverse topics in-the-realm of human-centered AI. Check out some of our recent works.

Authors: Tarun Jain, Yufei Gao, Sridhar Vanga, Karan Singla

We present a fine-tuned LLM framework for verifiable T.V. news QA pairs, surpassing similar base models and enhancing answer contextualization.

Venue: Under review at ICASSP 2025

Published on: Fri Jul 01 2022

Authors: Karan Singla, Shahab Jalavand, Andrej Ljolje, Antonio Moreno Daniel, Srinivas Bangalore, Yeon-Jun Kim, Ben Stern

1SPU extends ASR with tagged placeholders for semantic events, achieving improved transcription quality on SLUE and SLURP benchmarks.

Venue: International Conference on Natural Language Processing 2023

Published on: Fri Jul 01 2022

Authors: Karan Singla, Yeon-Jun Kim, Srinivas Bangalore

This study refines entity extraction directly from speech, optimizing ASR encoders to transcribe only relevant content in virtual agent dialogs.

Venue: Empirical Methods in Natural Language Processing 2023

Published on: Thu Sep 15 2022

Authors: Karan Singla, Mahnoosh Mehrabani, Daniel Pressel, Ryan Price, Bhargav S. Chinnari, Yeon-Jun Kim, Srinivas Bangalore

We introduce a multi-modal model for token-level classification using cross-modal attention, efficient for single GPU training.

Venue: International Conference on Natural Language Processing 2023

Published on: Thu Sep 15 2022

Authors: Yash Khasbage, Daniel Alcaraz Carrión, Jennifer Hinnell, Frankie Robertson, Karan Singla, Peter Uhrig, Mark Turner

The Red Hen Anonymizer enables de-identification of audiovisual data, ensuring privacy while supporting machine learning and research.

Venue: Linguistics Vanguard 2022 (Journal)

Published on: Thu Sep 15 2022