most citedCUE Vectors: Modular Training of Language Models Conditioned on Diverse Contextual Signals

2 citations · 2 across the 2 of their papers we have counts for

collaborators

2 papers

cs.CL2022★ 2 cited

CUE Vectors: Modular Training of Language Models Conditioned on Diverse Contextual Signals

Scott Novotney, Sreeparna Mukherjee, Zeeshan Ahmed +1

We propose a framework to modularize the training of neural language models that use diverse forms of sentence-external context (including metadata) by eliminating the need to join…

eess.AS2022

Mitigating Closed-model Adversarial Examples with Bayesian Neural Modeling for Enhanced End-to-End Speech Recognition

Chao-Han Huck Yang, Zeeshan Ahmed, Yile Gu +5

In this work, we aim to enhance the system robustness of end-to-end automatic speech recognition (ASR) against adversarially-noisy speech examples. We focus on a rigorous and empir…