KuralHub lists 4 speech emotion recognition datasets for Malayalam, 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.
IIITSaint-EmoMDB is a meticulously curated Malayalam speech corpus designed to explore the relationship between emotional perception and mental health conditions, such as depression and anxiety. The dataset includes speech samples annotated with emotion ratings on valence and arousal scales, alongside self-reported depression ratings. This resource aims to facilitate research in understanding the intricate connections between emotions and mental health.
Citation
@INPROCEEDINGS{10800549,
author={Thomas, Christa and Vats, Guneesh and Johnson, Aravind and George, Ashin and George, Talit Sara and Cherian, Reni K and Srivastava, Priyanka and Yarra, Chiranjeevi},
booktitle={2024 27th Conference of the Oriental COCOSDA International Committee for the Co-ordination and Standardisation of Speech Databases and Assessment Techniques (O-COCOSDA)},
title={IIITSaint-EmoMDB: Carefully Curated Malayalam Speech Corpus with Emotion and Self-Reported Depression Ratings},
year={2024},
volume={},
number={},
pages={1-6},
keywords={Databases;Anxiety disorders;Focusing;Mental health;Depression;emotional perception;depression;emotion;mental health;valence and arousal},
doi={10.1109/O-COCOSDA64382.2024.10800549}
}
The KMSER Dataset was created to address the lack of speech emotion recognition resources for the Malayalam language, specifically the Kannur dialect. Audio samples were collected from freely accessible YouTube videos, WhatsApp chats, and self-recorded files. The dataset focuses on four primary emotions: happy, angry, sad, and neutral. Feature extraction techniques such as Mel-Frequency Cepstral Coefficient (MFCC), chroma shift, spectral centroid, spectral bandwidth, rolloff, zero crossing rate, and mel spectrogram were employed. Classification was performed using Support Vector Machine (SVM) and Random Forest algorithms. (thegrenze.com)
Note: Access to the dataset may require a formal request. Please contact the provided email for more information.
Citation
@article{amitha2023kannur,
title={Malayalam Kannur Dialect Speech Emotion Recognition using Machine Learning},
author={Amitha, IC and Surendran, Amal and Likhith, NV and Adithya, PP and Aswin, K},
journal={Grenze International Journal of Engineering and Technology},
volume={9},
number={2},
year={2023}
}
MAVES DB is a speech emotion dataset created by Shabina Bhaskar in 2022. It is designed for emotion recognition tasks and contains audio samples with various emotional tones.
Citation
@misc{bhaskar2022maves,
author = {Bhaskar, Shabina},
title = {MAVES DB},
year = {2022},
publisher = {Mendeley Data},
version = {V1},
doi = {10.17632/6hm2xt5hd8.1}
}
This dataset is part of research focused on emotion recognition from speech using artificial neural networks. It includes gender-dependent databases and utilizes features like discrete wavelet transforms, low-pass filters, and frequency filtering to extract speech features for emotion classification. The dataset is part of an effort to advance affective computing and automatic emotion recognition.
Citation
@INPROCEEDINGS{5376782,
author={Firoz, S.A. and Raji, S.A. and Babu, A.P.},
booktitle={2009 International Conference on Advances in Computing, Control, and Telecommunication Technologies},
title={Automatic Emotion Recognition from Speech Using Artificial Neural Networks with Gender-Dependent Databases},
year={2009},
volume={},
number={},
pages={162-164},
keywords={Emotion recognition;Speech;Artificial neural networks;Databases;Discrete wavelet transforms;Low pass filters;Frequency;Filter bank;Filtering;Humans;Automatic Emotion Recognition;Affective Computing;Discrete Wavelet Transform;Artificial Neural Networks;Multi Layer Perceptron},
doi={10.1109/ACT.2009.49}}
Know another Malayalam speech emotion dataset, or a newer version of one listed here? Suggest it or open a pull request.
This page is generated from datasets/Malayalam/ in the KuralHub repository.
If you use a dataset, cite its authors. If KuralHub helped you find it, please also cite the KuralHub paper.