Man-Wai Mak

Emotion preference models can be dramatically improved by addressing both data sparsity and model bias, leading to more accurate emotional assessments in multimodal contexts.

Replacing standard residual connections with manifold-constrained hyper-connections boosts speaker recognition performance across multiple architectures.

Emotional inertia can significantly enhance emotion recognition accuracy in conversation, leading to a leap in performance over traditional methods.

Speaker recognition accuracy improves dramatically when leveraging a U-Net-based fusion of noisy and enhanced speech, coupled with a novel training strategy.