Alibaba Cloud Developer
Dec 25, 2018 · Artificial Intelligence
How a Composite Framework Boosts Speech Emotion Recognition in Noisy Environments
This paper presents a multi‑subsystem ensemble for voice‑based emotion recognition that leverages low‑level descriptors, high‑level iVector features, attention‑based RNNs, and text SVMs, achieving superior robustness and accuracy on the noisy MEC 2017 dataset.
attention modeldeep neural networkmultitask learning
0 likes · 8 min read