BIM-SSD-V1 Dataset (Malaysian Sign Language)

This dataset was developed for Malaysian Sign Language (Bahasa Isyarat Malaysia, BIM) to support research in sign language recognition and translation. Image and video data were collected from four BIM signers in both controlled and uncontrolled environments. There is a total of 4,858 video samples recorded using smartphone cameras. The signers performed a set of tasks that cover alphabets, numbers, words, and sentences, as outlined in the Sign Language Module for Dataset Development.pdf. All recordings were annotated with gloss labels verified by a BIM language expert. The videos were processed into sequential RGB frames and resized uniformly. These frames and their corresponding annotation files are compiled as BIM-SSD-V1. BIM-SSD-V1 dataset has been split into 4,458 train, 200 validation and 200 test sets. This dataset provides the first standardized and covered diverse communications used in various real-time scenarios, such as hospital, school and emergencies.