Salat Posture Recognition Dataset by Belkis Hassani Computer Vision Project

This dataset is designed for training and evaluating deep learning models for recognizing the various postures (steps) of Muslim prayer (salat). Each image was meticulously labeled with bounding boxes encompassing the individual performing the specific salat posture. This dataset involved: Selecting images: A few pictures were taken from different places to show various moves that comprise salat positions (1). Internet augmentation: More pictures showing people praying were added to the data set from the internet. The seven classes correspond to the essential positions in Islamic prayer:

1. raising: Standing posture.
2. ruku: Bowing posture.
3. sitting: Sitting position during prayer.
4. sujud: Prostration posture.
5. takbeer: Hands raised.
6. tashhud: Sitting posture for final supplications.
7. tasleem: Final salutation posture. Reference Koubâa, A., Ammar, A., Benjdira, B., Al-Hadid, A., Kawaf, B., Al-Yahri, S.A., Babiker, A., Assaf, K. and Ras, M.B., 2020, March. Activity Monitoring of Islamic Prayer (Salat) Postures using Deep Learning. In 2020 6th Conference on Data Science and Machine Learning Applications (CDMA) (pp. 106-111). IEEE.