Speeding-up normalized neural networks for face/object detection

  • Authors:
  • Hazem M. El-Bakry;Qiangfu Zhao

  • Affiliations:
  • University of Aizu, Aizu Wakamatsu, Japan;University of Aizu, Aizu Wakamatsu, Japan

  • Venue:
  • Machine Graphics & Vision International Journal
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

Finding an object or a face in an input image is a search problem in the spatial domain. Neural networks have shown good results in detecting a certain face/object in a given image. In this paper, faster neural networks for face/object detection are presented. Such networks are designed based on cross correlation in the frequency domain between the input image and the input weights of neural networks. This approach is developed to reduce the computation steps required by these faster neural networks for the search process. The principle of divide and conquer strategy is applied through image decomposition. Each image is divided into small-size sub-images, and then each of them is tested separately using a single faster neural network. Furthermore, the fastest face/object detection is achieved using parallel processing techniques to test the resulting sub-images simultaneously using the same number of taster neural networks. In contrast to using faster neural networks only, the speed-up ratio is increased with the size of the input image when using faster neural networks and image decomposition. Moreover, the problem of local subimage normalization in the frequency domain is solved. The effect of image normalization on the speed-up ratio for face/object detection is discussed. Simulation results show that local subimage normalization through weight normalization is faster than subimage normalization in the spatial domain. The overall speed-up ratio of the detection process is increased as the normalization of weights is carried out off line.