Incremental Convolution Neural Network and Its Application in Face Detection
Peng Hong-jing · Jisuanji fangzhen · 2009
Convolution neural networks have the capability of extracting hidden discriminating features,so that they successfully have been applied in the fields of face detection,etc.Due to their fixed architecture,the scale of neural networks usually depends on the designer's experience and prevents them from learning further.A novel approach of constructing CNN with varying architecture was proposed.Training neural networks started off with the simplest architecture with single neuron each layer,then adding new neural cells in every layer and modifying their corresponding weights continuously,until training target was reached.In the face detection experiments,various results with different network scales were given.These results illustrate that the CNN can get a good tradeoff between the detection rate and the architecture size.Moreover,if appending new complementing samples,the detection rate can be improved significantly,only adjusting a few adding neurons' weights while keeping previous weights unchanging.