Bindu Garg

Dr. Bindu Garg

orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 January 2025 in International Journal of Engineering
Verified
Journal Article
Open access
PDF file
MKMANISHA Kasar
MANISHA Kasar
PKP. Kavimandan
P. Kavimandan
TST. Suryawanshi
T. Suryawanshi
BGBindu Garg
Bindu Garg
Facial emotions are prime characteristics of humans that reflect the attributes of a human emotional condition. Facial Emotional Recognition (FER) is a significant research area since it has several significant applications such as crime detection, security, government sectors, surveillance, etc. Many researchers have addressed FER and put efforts into improving recognition using various methods such as Deep Convolutional Neural Networks (DCNN), Local Binary Pattern Convolutional Neural Networks (LBPCNN) and Micro Expression Recognition (MER). Nevertheless, there is a dearth of better Convolutional Neural Networks (CNN) for better accuracy of FER. Various methods have already proven their accuracies on datasets such as Fer_2013, CK48 and Legend. There are many challenges such as varying positions of images, illumination and noise etc. are not resolved yet. In this direction, the proposed modified CNN model has a combination of CNN layers to train the model and achieved better training and validation accuracy on datasets. The proposed method has three-fold contributions (1) to propose an efficient CNN model for FER and (2) to train the proposed model on three standard datasets 3) to test on real-time images. The proposed model is trained on three standard datasets (Fer_2013, CK48, and Legend) and achieves significantly higher training accuracy 92.34%, 99.84%, and 89.86%, respectively as compared to previous methods. Furthermore, the model demonstrates impressive generalization ability by achieving even higher test accuracy on real-world images 91.54%, 96.27%, and 87.47%, respectively. Also, the proposed novel CNN model has addressed all current challenges of FER with low computational complexity and prediction of facial expression more efficiently.

21 November 2023 in Multimedia Tools and Applications
Verified
Journal Article
Closed access
PDF file
AGAnushree Goud
Anushree Goud
BGBindu Garg
Bindu Garg

11 March 2022
Verified
Book Chapter
Closed access
GCGyankamal J. Chhajed
Gyankamal J. Chhajed
BGBindu Garg
Bindu Garg
In today's digital world, it is noticeable that images are transferred in binary form which carries significant and confidential information. To secure and keep such information unnoticeable is the prime concern during transmission. Information hiding as watermarking or secret messaging are most acceptable ways to attain this. In this direction, a new information flow system using data hiding approach is proposed in this chapter. A decision tree is applied for locating the most suitable blocks for data hiding in the presented technique. Decision trees are one of the best techniques used as a classifier. In this approach, the decision tree classifies a particular block to be selected for hiding information, based on the LURD pattern of 3X3 pixel block of image and the blocks of information to be hidden. For security, information is encrypted and then scattered in the image. The receiver can retrieve the data without the original image. The results prove that the proposed scheme is capable of minimizing distortion caused due to data hiding.


27 May 2025 in Data in Brief
Verified
Journal Article
Open access
BGBindu Garg
Bindu Garg
MKMANISHA Kasar
MANISHA Kasar
PPPriyanka Paygude
Priyanka Paygude
ADAmol Dhumane
Amol Dhumane
SASrinivas Ambala
Srinivas Ambala
AKAchyut Kashyap
Achyut Kashyap
Sign language is a very important mode of communication among deaf and hard-of-hearing populations. Automatic sign language detection based on deep learning model is the theme of this study. Hand gestures are classified by the Convolutional Neural Network (CNN) model to different signs. For training purposes, there are 26,000 images available with 3000 images for every alphabet letter such that there is complete representation of sign language gesture. Photos were taken in controlled lighting with a consistent black background to facilitate better feature extraction. The data contains varied participants of various age groups, skin types, and hand shapes to enhance generalization. Data collection was standardized through iPhone 15 Pro Max, black background cloth, tripod stand, and remote-controlled Drodcam app to maintain consistency in image quality and framing. For diversity and realism, three participants were involved in data collection, each providing 1000 images per sign, resulting in a rich and diverse dataset. Preprocessing of data methods were used for achieving the best quality of data, such as resizing, conversion to grayscale, normalization, and augmentation. Different techniques of data augmentation like rotation, flipping, scaling, brightness change, and addition of Gaussian noise were used to introduce variations in hand gestures and make the model robust against various environmental conditions. The dataset was then partitioned into 70 % training, 15 % validation, and 15 % test sets for maximizing model performance and ensuring good generalization. The dataset show high accuracy, reflecting the potential of the model for real-world usage, such as accessibility tools for the deaf community, educational tools, and real-time sign language recognition systems.