Skip to main content

Table 13 SER on the Albayzín 2012 test partition for different systems proposed in the literature compared to our proposed RNN approach

From: Multiclass audio segmentation based on recurrent neural networks for broadcast domain data

SystemSER
RNN proposal (pool + mixup)24.93
DCASE baseline31.21
GMM + Viterbi decoding [67]26.34
HMM-GMM [68]26.53