Ego-SLD: A Video Dataset of Egocentric Action Recognition for Bengali Sign Language Detection

Citation Author(s):
Asfak
Ali
Department of Electronics and Telecommunication Engineering, Jadavpur University, Kolkata, India
Bibek
Das
Department of Electronics and Telecommunication Engineering, Jadavpur University, Kolkata, India
Ahmad Sami
Al-Shamayleh
Department of Data Science and Artificial Intelligence, Faculty of Information Technology, Al-Ahliyya Amman University, Amman, 19328, Jordan
Sujatro
Sarkar
Department of Electrical Engineering, Jadavpur University, Kolkata, India
Chinmoyee
Ray
Department of Electronics and Telecommunication Engineering, Jadavpur University, Kolkata, India
Sayoni
Mandal
Department of Electronics and Telecommunication Engineering, Jadavpur University, Kolkata, India
Saifuddin
Sk
Department of Electronics and Telecommunication Engineering, Jadavpur University, Kolkata, India
Adnan
Akhunzada
College of Computing and Information Technology, University of Doha for Science and Technology, Doha, Qatar
Submitted by:
Asfak Ali
Last updated:
Sun, 01/19/2025 - 13:19
DOI:
10.21227/kbeg-fa06
Data Format:
Video
License:
Creative Commons Attribution
131 Views
Categories:
Artificial Intelligence
Machine Learning
Computer Vision
Keywords:
Sign Language Detection, Video Action Recognition, hand gesture recognition, and Video Data
0
0 ratings - Please login to submit your rating.
ACCESS DATASET CITE SHARE/EMBED

Abstract 

Sign Language Recognition integrates computer vision and natural language processing to automatically interpret hand gestures and translate them into spoken or written Bengali. The primary goal is to bridge the communication gap between sign language users and non-users by recognizing gestures, movements, postures, and facial expressions that correspond to spoken language elements. Since hand gestures are the cornerstone of sign language communication, they play a pivotal role in improving the accuracy of sign language recognition systems. This article introduces Ego-SLD, a video dataset specifically designed for Egocentric Action Recognition of Bangla Sign Language used in daily life situations. The dataset contains videos of 16 commonly used words, collected from 12 individuals (10 males and 2 females) aged between 16 and 42 years. Each participant provided two samples of each word in an indoor environment with standard lighting conditions. Ego-SLD is a crucial resource for the automatic recognition of Bangla sign language and holds significant potential to benefit the deaf community. Moreover, it serves as a valuable dataset for researchers focusing on vision-based sign language detection, hand gesture recognition, and egocentric action recognition.

Instructions: 

The dataset comprises videos organized into 16 folders, each corresponding to one of the 16 classes, named accordingly. No specialized software is required to use this dataset. The dataset is systematically organized into 16 distinct sign gesture classes: “Thanks,” “Support,” “Sorry,” “Sewing,” “Secrecy,” “Question,” “Put off the Light,” “Please,” “Meet,” “I Love You,” “Hello,” “Goodbye,” “Deaf,” “Dangerous,” “Complain,” and “Acting.”. The inclusion of diverse gender representations, varied backgrounds, and different hand usage aims to enhance the dataset's potential to improve generalization performance in machine learning classification tasks.

 

Dataset Files

LOGIN TO ACCESS DATASET FILES

 

QUESTIONS?