Two papers are accepted in IEEE iSES 2021.

1.Rakesh Reddy Yakkati, Sreenivasa Reddy Yeduri, Linga Reddy Cenkeramaddi, “Hand Gesture Classification Using Grayscale Thermal Images and Convolution Neural Network,” has been accepted in IEEE International Symposium on Smart Electronic Systems, 2021.

Keywords:Convolution neural network, image classification, hand gesture, classification accuracy, and inference time.

Abstract:In this paper, we propose a convolution neural network for classifying grayscale images of hand gestures. For classification, we look at ten different hand gestures collected
from various people using a thermal camera. We then compare the proposed model’s performance in terms of classification accuracy and inference time to that of other benchmark models.We demonstrate through extensive results that the proposed
model achieves higher classification accuracy while using a smaller model size. Furthermore, we show that the proposed model outperforms benchmark models in terms of inference time.

2.Muralidhar Reddy Challa, Abhinav Kumar and Linga Reddy Cenkeramaddi, “Face Recognition Using mmWave RADAR Imaging,” has been accepted in IEEE International Symposium on Smart Electronic Systems, 2021.

Keywords:Machine learning algorithms,Convolution,Face recognition, Signal processing algorithms,Radar imaging,Feature extraction,Power systems

Abstract:The current work presents a novel approach to signal processing and face recognition based on 60 GHz mmWave RADAR imaging. Machine learning algorithms such as Convolutional Auto-Encoder and Random Forest algorithm are employed to implement the face recognition scheme. The work presents an approach towards computing and processing higher dimensional RADAR imaging information through extreme feature extraction followed by simple Random Forest, thus enabling a computationally inexpensive algorithm for a mobile friendly implementation.

More details:DOI: 10.1109/iSES52644.2021.00081

Leave a Reply

Your email address will not be published. Required fields are marked *