Qualitative Study of Gesture Annotation Corpus: Challenges and Perspectives
Keywords: Gesture corpora, Enrichment, Annotation, Lexical affiliates, Gesture labelling
TL;DR: This paper critically reviews corpus enrichment in the BEAT dataset, focusing on gesture annotation challenges and proposing strategies to advance gesture research methodologies.
Abstract: Effective data management and corpus enrichment are essential for advancing research methodologies in gesture studies. This paper critically examines the practices surrounding data management and corpora enrichment within a gesture dataset, focusing on qualitative analysis and methodological challenges. It identifies key issues in gesture annotation, including segmentation, labeling gestures, and lexical affiliates, revealing significant discrepancies and highlighting the complexities in interpretation. Despite these challenges, the inclusion of gesture dataset annotations marks progress in gesture research, offering opportunities for refining methodologies and enhancing data utilization. Strategies proposed aim to improve annotation practices, promote methodological transparency, and ensure the reliability of enriched corpora for nuanced analysis in gesture studies. This study contributes to advancing gesture research methodologies, emphasizing the importance of rigorous annotation protocols and fostering a standardized approach to enhance the utility and reliability of annotated datasets.
Submission Number: 4
Loading