Object Detection Using Deep Learning

Authors

  • H. Harshita Kumar UG Student, Department of Computer Science and Engineering, Sri Venkateshwara College of Engineering, Bangalore, India
  • Harshitha Manjunath UG Student, Department of Computer Science and Engineering, Sri Venkateshwara College of Engineering, Bangalore, India
  • Chandana Shivanna UG Student, Department of Computer Science and Engineering, Sri Venkateshwara College of Engineering, Bangalore, India
  • Mangala Manjunath UG Student, Department of Computer Science and Engineering, Sri Venkateshwara College of Engineering, Bangalore, India

Keywords:

Activation function, Bounding Box, Convolution, Deep Learning, Feature map, Object detection, Pooling, Softmax, Testing, Training

Abstract

This paper proposes an efficient and an accurate object detection system using Deep Learning [3]. Compared to the existing systems, it makes sure it deduces the drawbacks and provides a better solution. In a certain image as an input, the system divides the image into bounding boxes [7] and associates the class probabilities. In one evaluation, a single neural network predicts the bounding boxes and its probabilities. These systems use classifier for a specific object in an image and evaluate it at every location, corner of the image. Once the classification is done, it makes sure there are no duplicate detections and labels the objects with its accuracy score. Since our system uses deep learning, it makes use of Convolutional Neural Network algorithm to compute the images in a more perfect manner. Our architecture is less complicated and thus makes it fast to process.

Downloads

Download data is not yet available.

Downloads

Published

04-08-2020

How to Cite

[1]
H. H. Kumar, H. Manjunath, C. Shivanna, and M. Manjunath, “Object Detection Using Deep Learning”, IJRESM, vol. 3, no. 8, pp. 36–39, Aug. 2020.

Issue

Section

Articles