The Multimodal Video Search by Examples (MVSE) project investigates usin...
Prompt-based classifiers are an attractive approach for zero-shot
classi...
Evaluating Natural Language Generation (NLG) outputs is crucial but labo...
A crucial part of an accurate and reliable spoken language assessment sy...
Multiple Choice examinations are a ubiquitous form of assessment that is...
As speech recognition model sizes and training data requirements grow, i...
State space models (SSMs) have recently shown promising results on
small...
State-of-the-art sequence-to-sequence models often require autoregressiv...
Generative Large Language Models (LLMs) such as GPT-3 are capable of
gen...
Error correction models form an important part of Automatic Speech
Recog...
State-of-the-art summarization systems can generate highly fluent summar...
There has been a growing demand for automated spoken language assessment...
Automatic summary assessment is useful for both machine-generated and
hu...
Distributional shift, or the mismatch between training and deployment da...
Deep learning is increasingly being applied in safety-critical domains. ...
Transformer models have achieved state-of-the-art results in a wide rang...
There has been significant research done on developing methods for impro...
Text-based machine comprehension (MC) systems have a wide-range of
appli...
Transformer-based models have achieved state-of-the-art results in a wid...
Auto-regressive sequence-to-sequence models with attention mechanisms ha...
Automatic spoken language assessment systems are becoming more popular i...
Auto-regressive sequence-to-sequence models with attention mechanism hav...