This repository contains a refined version of the Respiratory Sound Database, collected from Kaggle. The dataset includes respiratory sound recordings, which can be used for machine learning models aimed at detecting lung conditions like wheezes and crackles.
- High-quality respiratory sound recordings.
- Suitable for research and machine learning applications related to respiratory health.
- Preprocessed data for efficient use in model training and testing.
This repository provides the dataset in a well-organized structure. To use the dataset, simply download the files and use them for your machine learning or signal processing tasks.
- /audio: Contains the audio files of respiratory sounds.
- /annotations: Includes the annotations for each audio file, such as the presence of crackles and wheezes.
- /metadata: Metadata about each subject, including age, gender, and diagnosis.
This dataset is a refined form of the original data collected from Kaggle.