Indian Sign Language Skeletal-point NumPy array using MediaPipe

Citation Author(s):
Jans
Johnson
Jisha
Joseph
Maris
Reji
Megha
George
Submitted by:
Jisha Joseph
Last updated:
Mon, 04/24/2023 - 15:53
DOI:
10.21227/6emn-2r12
Data Format:
*.npy
Links:
Indian Sign Language Detector
License:
Creative Commons Attribution
1249 Views
Categories:
Machine Learning
Computer Vision
Keywords:
Indian Sign Language, sign language alphabet, MediaPipe, NumPy arrays
3.5
2 ratings - Please login to submit your rating.
ACCESS DATASET CITE SHARE/EMBED

Abstract 

The dataset consists of NumPy arrays for each alphabet in Indian Sign Language, excluding 'R'. The NumPy arrays denote the (x,y,z) coordinates of the skeletal points of the left and right hand (21 skeletal points each) for each alphabet. Each alphabet has 120 sequences, split into 30 frames each, giving 3600 .np files per alphabet, using MediaPipe.

The dataset is created on the basis of skeletal-point action recognition and key-point collection.

Instructions: 

The NumPy arrays can be used as input data into various ML models to train it in recognizing alphabets in Indian Sign Language

Comments

really helpful

Submitted by Kanika Jadhav on Thu, 08/22/2024 - 12:36

really helpful for my project

Submitted by Sahil Rajwar on Sun, 09/08/2024 - 07:53

Dataset Files

LOGIN TO ACCESS DATASET FILES

Documentation

AttachmentSize
File NumPy arrays of alphabets in ISL except 'R' using skeletal-point recognition246.99 KB

 

QUESTIONS?