Jun 2026
LOPA: Enhancing Spoken Language Assessment via Latent Ordinal Prototype Alignment
This paper works around the necessity of large-scale MLLMs by introducing Latent Ordinal Prototype Alignment (LOPA) for SLA, a prototype-based regularizer that enforces an ordinal geometric prior directly on the latent space.
Hong-Yun Lin, Fu-An Chao, Bi-Cheng Yan et al.
· arXiv.org · 0 citations