Large Raw Emotional Dataset with Aggregation Mechanism

We present a new data set for speech emotion recognition (SER) tasks called Dusha. The corpus contains approximately 350 hours of data, more than 300 000 audio recordings with Russian speech and their transcripts. Therefore it is the biggest open bi-modal data collection for SER task nowadays. It is annotated using a crowd-sourcing platform and includes two subsets: acted and real-life. Acted subset has a more balanced class distribution than the unbalanced real-life part consisting of audio podcasts. So the first one is suitable for model pre-training, and the second is elaborated for fine-tuning purposes, model approbation, and validation. This paper describes pre-processing routine, annotation, and experiment with a baseline model to demonstrate some actual metrics which could be obtained with the Dusha data set.

PDF Abstract


Introduced in the Paper:


Used in the Paper:

Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Speech Emotion Recognition Dusha Crowd Dusha baseline UA 0.83 # 1
WA 0.76 # 1
Macro F1 0.77 # 1
Speech Emotion Recognition Dusha Podcast Dusha baseline UA 0.89 # 1
WA 0.53 # 1
Macro F1 0.54 # 1


No methods listed for this paper. Add relevant methods here