datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
modified_shemo
Dataset Card for Modified SHEMO
Dataset Summary
This dataset is a corrected and modified version of the Sharif Emotional Speech Database (ShEMO) named modified_shemo. The original dataset contained significant mismatches between audio files and their corresponding transcriptions. This version resolves those issues, resulting in a cleaner and more reliable resource for Persian Speech Emotion Recognition (SER) and Automatic Speech Recognition (ASR).
Curation and… See the full description on the dataset page: https://huggingface.co/datasets/aliyzd95/modified_shemo.modicol
MoDiCoL - A Modular Diagnostic Continual Learning Dataset for ASR
MoDiCoL is a speech dataset designed to study the robustness of ASR models to different drift factors in a controlled, continual setting. We construct MoDiCoL using a systematic factorial design that enables a rigorous evaluation of linguistic, speaker, and acoustic variation with clearly defined experimental runs. By combining real-world and synthetic speech with a configuration-dependent augmentation pipeline… See the full description on the dataset page: https://huggingface.co/datasets/TPekarekRosin/modicol.