A Dimensional Approach to Canine Bark Analysis for Assistance Dog Signaling

Read full text Summarise with VIRAT
Authors: Hailin Song, Shelley Brady, Tomas Emmanuel Ward, Alan F. Smeaton
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Standard classification of canine vocalizations is infeasible for assistance dogs, where data is inherently sparse and ethically constrained. We reframe this problem as a continuous regression task within a two-dimensional arousal-valence space. Central to our approach is an adjusted Siamese Network trained not on binary similarity, but on the ordinal and numeric distance between input pairs. Trained on the public Emotional Canines dataset, our model reduces critical "turn-around" errors by up to 50% on the challenging valence dimension compared to a regression baseline. Qualitative validation on real-world dataset confirms the learned space is semantically meaningful, establishing a proof-of-concept for robustly analyzing canine bark under severe data limitations.

Fields of Study & Subjects

Emotion and Mood Recognition Human-Animal Interaction Studies Music and Audio Processing

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
22 March 2026
Language
EN