116 followers
Sparse MERIT, a mixture-of-experts framework, improves Speech Emotion Recognition and Speech Enhancement jointly; outperforms baselines by 12% F1-macro in noisy conditions on the MSP-Podcast corpus.
So far, Altmetric has seen
3 Bluesky posts from 3 Bluesky users, with an upper bound of 423 followers.
263 followers
RP @eessas-bot.bsky.social: Jing-Tong Tzeng, Carlos Busso, Chi-Chun Lee: Joint Learning using Mixture-of-Expert-Based Representation for Enhanced Speech Generation and Robust Emotion Recognition https://arxiv.org/abs/2509.08470 https://arxiv.org/pdf/250…
44 followers
Jing-Tong Tzeng, Carlos Busso, Chi-Chun Lee: Joint Learning using Mixture-of-Expert-Based Representation for Enhanced Speech Generation and Robust Emotion Recognition https://arxiv.org/abs/2509.08470 https://arxiv.org/pdf/2509.08470 https://arxiv.org/ht…