# ShEMO -- A Large-Scale Validated Database for Persian Speech Emotion Detection

## TL;DR

This paper presents ShEMO, a large validated Persian speech emotion database with 3000 utterances, and provides benchmark results using common classification methods to facilitate future research in Persian speech emotion detection.

## Contribution

Introduction of the first large-scale validated Persian speech emotion database with benchmark results for emotion detection.

## Key findings

- Support vector machine achieved 58.2% accuracy in gender-independent models.
- Gender-dependent models achieved around 57.6-59.4% accuracy.
- Inter-annotator agreement was substantial at 64%.

## Abstract

This paper introduces a large-scale, validated database for Persian called Sharif Emotional Speech Database (ShEMO). The database includes 3000 semi-natural utterances, equivalent to 3 hours and 25 minutes of speech data extracted from online radio plays. The ShEMO covers speech samples of 87 native-Persian speakers for five basic emotions including anger, fear, happiness, sadness and surprise, as well as neutral state. Twelve annotators label the underlying emotional state of utterances and majority voting is used to decide on the final labels. According to the kappa measure, the inter-annotator agreement is 64% which is interpreted as "substantial agreement". We also present benchmark results based on common classification methods in speech emotion detection task. According to the experiments, support vector machine achieves the best results for both gender-independent (58.2%) and gender-dependent models (female=59.4%, male=57.6%). The ShEMO is available for academic purposes free of charge to provide a baseline for further research on Persian emotional speech.