Fetching the paper…

Cascaded Cross-Module Residual Learning towards Lightweight End-to-End Speech Coding · Around