Resource Evaluation for Usable Speech Interfaces: Utilizing Human-Human Dialogue
Human-human spoken dialogues are considered an important tool for effective speech interface design and are often used for stochastic model training in speech based applications. However, the less restricted nature of human-human interaction compared to human-system interaction may undermine the usefulness of such corpora for creating effective and usable interfaces. In this respect, this work examines the differences between corpora collected from human-human interaction and corpora collected from actual system use, in order to formally assess the appropriateness of the former for both the design and implementation of spoken dialogue systems. Comparison results show that there are significant differences with respect to vocabulary, sentence structure and speech recognition success rate among others. Nevertheless, compared to other available tools and techniques, human-human dialogues may still be used as a temporary at least solution for building more effective working systems. Accordingly, ways to better utilize such resources are presented.
Code (0)
등록된 구현이 없습니다.
Tasks
Dialogue ManagementLanguage ModellingSentencespeech-recognitionSpeech RecognitionSpoken Dialogue SystemsText GenerationSimilar Papers 제목 키워드 기반
Working with Trouble and Failures in Conversation between Humans and Robots (WTF 2023) & Is CUI Design Ready Yet?
Workshop proceedings of two co-located workshops "Working with Troubles and Failures in Conversation with Humans and Robots" (WTF 2023) and "Is CUI Design Ready Yet?", both of which were part of the ACM conference on con…
NLP Lean Programming Framework: Developing NLP Applications More Effectively
This paper presents NLP Lean Programming framework (NLPf), a new framework for creating custom Natural Language Processing (NLP) models and pipelines by utilizing common software development build systems. This approach …
Named Entity Recognition (NER)Transfer Learning Based Free-Form Speech Command Classification for Low-Resource Languages
Current state-of-the-art speech-based user interfaces use data intense methodologies to recognize free-form speech commands. However, this is not viable for low-resource languages, which lack speech data. This restricts …
FormGeneral ClassificationTransfer LearningLessons from Usable ML Deployments and Application to Wind Turbine Monitoring
Through past experiences deploying what we call usable ML (one step beyond explainable ML, including both explanations and other augmenting information) to real-world domains, we have learned three key lessons. First, ma…
Decision MakingLarge Language Model Based Generative Error Correction: A Challenge and Baselines for Speech Recognition, Speaker Tagging, and Emotion Recognition
Given recent advances in generative AI technology, a key question is how large language models (LLMs) can enhance acoustic modeling tasks using text decoding results from a frozen, pretrained automatic speech recognition…
Automatic Speech RecognitionAutomatic Speech Recognition (ASR)Emotion RecognitionLanguage Modeling+4