Reference-Aware SFM Layers for Intrusive Intelligibility Prediction

Intrusive speech-intelligibility predictors that exploit explicit reference signals are now widespread, yet they have not consistently surpassed non-intrusive systems. We argue that a primary cause is the limited exploitation of speech foundation models (SFMs). This work revisits intrusive prediction by combining reference conditioning with multi-layer SFM representations. Our final system achieves a root mean squared error (RMSE) 22.36 on the development set and 24.98 on the evaluation set, ranking 1st on CPC3. These findings provide practical guidance for constructing SFM-based intrusive intelligibility predictors.