KuralHub lists 3 speech emotion recognition datasets for Burmese, 1 of them openly available. Each entry below gives the emotion labels, size, recording setup, licence, where to get it and how to cite it.
The Burmese Movies Interviews Speech Emotion Corpus (BMISEC) is a collection of natural speech segments extracted from Myanmar movies and interview recordings. The dataset consists of 1,332 utterances covering seven primary emotions. It is designed for Speech Emotion Recognition (SER) research and has been validated using DenseNet-Emotion.
Citation
@inproceedings{mar2023bmisec,
title={BMISEC: Corpus Of Burmese Emotional Speech},
author={Mar, Lwin Lwin and Pa, Win Pa and Nwe, Tin Lay},
booktitle={2023 IEEE Conference on Computer Applications (ICCA)},
pages={248--253},
year={2023},
organization={IEEE}
}
The ESMBS (Emotional Speech of Mandarin and Burmese Speakers) dataset was collected for emotion recognition research. It consists of speech samples from 12 non-professional actors (six Mandarin and six Burmese speakers, with an equal gender distribution). Each speaker recorded ten utterances per emotion across six emotions: anger, disgust, fear, joy, sadness, and surprise, resulting in a total of 720 emotional utterances.
Key details:
Audio Format: PCM WAVE, 16-bit, 22,050 Hz
Average Sample Length: 1.50 seconds
Emotion Evaluation: Four listeners, who did not understand the languages, judged the emotional content based on vocal characteristics alone.
Human Classification Accuracy: 65.7% overall (68.3% for Burmese, 63.1% for Mandarin), aligning with previous cross-cultural emotion recognition studies (55%-70% accuracy).
Citation
@article{nwe2003speech,
title={Speech emotion recognition using hidden Markov models},
author={Nwe, Tin Lay and Foo, Say Wei and De Silva, Liyanage C},
journal={Speech communication},
volume={41},
number={4},
pages={603--623},
year={2003},
publisher={Elsevier}
}
This dataset is used for speech-based emotion classification in the Burmese language. It was introduced in the research conducted by Tin Lay Nwe, Foo Say Wei, and Liyanage C. De Silva, focusing on recognizing emotions from speech signals. The dataset is useful for Speech Emotion Recognition (SER) tasks.
Citation
@inproceedings{nwe2001speech,
title={Speech based emotion classification},
author={Nwe, Tin Lay and Wei, Foo Say and De Silva, Liyanage C},
booktitle={Proceedings of IEEE region 10 international conference on electrical and electronic technology. TENCON 2001 (Cat. No. 01CH37239)},
volume={1},
pages={297--301},
year={2001},
organization={IEEE}
}
Know another Burmese speech emotion dataset, or a newer version of one listed here? Suggest it or open a pull request.
This page is generated from datasets/Burmese/ in the KuralHub repository.
If you use a dataset, cite its authors. If KuralHub helped you find it, please also cite the KuralHub paper.