EFFICIENT GESTURE CLASSIFICATION WITH PRE-TRAINED NEURAL NETWORKS

Authors

  • Maile Vaneshri Tulsidas M.Tech Student, Department Of Computer Science and Engineering, St. Johns College Of Engineering & Technology, Yerrakota, Yemmiganur, Kurnool Author
  • Dr. Y. Narasimha Reddy Associate Professor, Department Of Computer Science and Engineering, St. Johns College Of Engineering & Technology, Yerrakota, Yemmiganur, Kurnool Author

Keywords:

Neural Networks (CNNs), Vision Transformers (ViTs),, Recurrent Neural Networks (RNNs),

Abstract

Gesture recognition plays a crucial role in 
human-computer interaction, enabling intuitive 
and contactless communication between users 
and machines. Traditional approaches to gesture 
classification often require extensive training 
data and computational resources, making real
time implementation challenging. This study 
explores the effectiveness of pre-trained neural 
networks for efficient gesture classification, 
leveraging transfer learning techniques to 
improve accuracy while reducing training time. 
By utilizing models such as Convolutional 
Neural Networks (CNNs), Vision Transformers 
(ViTs), and Recurrent Neural Networks (RNNs), 
the system extracts high-level spatial and 
temporal features from gesture datasets. The 
proposed approach enhances recognition speed, 
optimizes resource utilization, and ensures 
robustness across different lighting conditions 
and 
gesture classification span across various 
domains, including sign language interpretation, 
virtual reality (VR), augmented reality (AR), 
robotics, and smart home automation. Accurate 
and efficient gesture recognition systems can 
enhance accessibility, improve user experiences, 
and eliminate the need for physical input 
devices. 
Traditional gesture recognition techniques rely 
on 
handcrafted 
feature 
extraction 
and 
conventional machine learning models, which 
often struggle with variations in lighting, 
background noise, and user-specific differences. 
These methods require extensive preprocessing 
and large amounts of labeled training data, 
making real-time implementation challenging. 
Additionally, training deep learning models 
from scratch for gesture classification is 
computationally expensive and time-consuming, 
limiting 
their 
environments. 
demonstrate 
that 
Experimental 
pre-trained 
results 
models 
significantly outperform conventional methods 
in accuracy and generalization, making them 
ideal for real-world applications such as sign 
language 
interpretation, 
virtual reality interactions, and smart home automation. This 
research highlights the potential of transfer 
learning in gesture recognition, paving the way 
for more accessible and efficient human
machine interfaces. 

References

Sunitha K. A, Anitha Saraswathi.P, Aarthi.M, Jayapriya. K, Lingam Sunny, “Deaf Mute Communication Interpreter- A Review”, International Journal of Applied Engineering Research, Volume 11, pp 290- 296, 2016.

Mathavan Suresh Anand, Nagarajan Mohan Kumar, Angappan Kumaresan, “An Efficient Framework for Indian SignLanguage Recognition Using Wavelet Transform” Circuits and Systems, Volume 7, pp 1874- 1883, 2016.

Mandeep Kaur Ahuja, Amardeep Singh, “Hand Gesture Recognition Using PCA”, International Journal of Computer Science Engineering and Technology (IJCSET), Volume 5, Issue 7, pp. 267-27, July 2015

Sagar P. More, Prof. Abdul Sattar, “Hand gesture recognition system for dumb people” International Journal of Science and Research (IJSR)

Chandandeep Kaur, Nivit Gill, “An Automated System for Indian Sign Language Recognition”, International Journal of Advanced Research in Computer Science and Software Engineering.

Pratibha Pandey, Vinay Jain, “Hand Gesture Recognition for Sign Language Recognition: A Review”, International Journal of Science, Engineering and Technology Research (IJSETR), Volume 4, Issue 3, March 2015.

Neelam K. Gilorkar, Manisha M. Ingle, “Real Time Detection and Recognition of Indian and American Sign Language Using Sift”, International Journal of Electronics and Communication Engineering & Technology (IJECET), Volume 5, Issue 5, pp. 11-18, May 2014

Downloads

Published

2026-01-19