analiza aiqbench
This research introduces "prosodic ABX," a novel framework extending the standard ABX discrimination task to directly measure how self-supervised speech models capture prosodic features like stress, pitch accent, and tone. By utilizing minimal pairs without explicit labels, the study provides a practical method for evaluating model sensitivity across low-resource settings and multiple languages. The findings offer critical insights into the internal representations of state-of-the-art audio AI systems, specifically regarding their ability to distinguish non-phonemic linguistic contrasts.
Prosodic ABX: A Language-Agnostic Method for Measuring Prosodic Contrast in Speech Representations
arXiv:2604.02102v2 Announce Type: replace-cross Abstract: Speech representations from self-supervised speech models (S3Ms) are known to be sensitive to phonemic contrasts, but their sensitivity to prosodic contrasts has not been directly measured. The ABX discrimination task has been used to measure phonemic contrast in S3M representations via minimal pairs. We introduce prosodic ABX, an extension of this framework to evaluate prosodic contrast with only a handful of examples and no explicit labels. Also, we build and release a dataset of English and Japanese minimal pairs and use it along with a Mandarin dataset to evaluate contrast in English stress, Japanese pitch accent, and Mandarin tone. Finally, we show that model and layer rankings are often preserved across several experimental conditions, making it practical for low-resource settings.
Abstract:Speech representations from self-supervised speech models (S3Ms) are known to be sensitive to phonemic contrasts, but their sensitivity to prosodic contrasts has not been directly measured. The ABX discrimination task has been used to measure phonemic contrast in S3M representations via minimal pairs. We introduce prosodic ABX, an extension of this framework to evaluate prosodic contrast with only a handful of examples and no explicit labels. Also, we build and release a dataset of English and Japanese minimal pairs and use it along with a Mandarin dataset to evaluate contrast in English stress, Japanese pitch accent, and Mandarin tone. Finally, we show that model and layer rankings are often preserved across several experimental conditions, making it practical for low-resource settings.
Computation and Language (cs.CL); Machine Learning (cs.LG); Sound (cs.SD); Audio and Speech Processing (eess.AS)
arXiv:2604.02102 [cs.CL]
(or arXiv:2604.02102v2 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2604.02102 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 223 cuvinte
Subiecte conexe
Latent reasoning architectures would likely undermine CoT, our strongest oversight tool
AI Alignment Forum · 23 sept. 2026
[AINews] OpenAI shuts off Cursor
Latent Space · 29 aug. 2026
Quantifying Consonant Contributions to Word Intelligibility via Acoustic Masking
arXiv · 14 sept. 2026
Story Imprinting: AI Assistants Absorb Traits from Human Characters They Resemble
LessWrong · 21 sept. 2026