Preprint
Aug 2026
Beyond Residual Connections: Manifold-Constrained Hyper-Connections for Robust Speaker Representation Learning
Manifold-Constrained Hyper-Connections (mHC) is introduced, reformulat- ing residual paths as a multi-stream evolution where informa- tion is mixed through a doubly stochastic matrix, highlighting its effectiveness for robust speaker representation learning.
Zezhong Jin, Xiaoyu Wang, Zhe Li et al.
· 0 citations