Hitachi Astemo Ltd.

Daichi Murata

Goichi Ono

Akira Kitayama

Tadashi Kishimoto

When it is assumed that a large-scale Deep Neural Network for autonomous driving applied compression, there are problems of a decrease in recognition accuracy of a post-compression Neural Network (NN) model and an increase in a compression design period, due to a large number of harmful or unnecessary training images (invalid training images). A training image selection unit Bcalculates an influence value on an inference, and generates an indexed training image set-necessary for an NN compression design, by using the influence value. A neural network compression unit Pnotified of the result via a memory Pcompresses the NN.

Neural network compression device and method for same

An arithmetic operation device causes a convolution arithmetic unit to perform a convolution arithmetic operation between a filter and target data corresponding to a size of the filter in each of a plurality of convolution layers constituting a neural network. The arithmetic operation device includes: a bit reduction unit that reduces a bit string corresponding to a first bit number from a least significant bit of the target data and reduces a bit string corresponding to a second bit number from a least significant bit of a weight that is an element of the filter for each convolution layer; and a bit addition unit that adds a bit string corresponding to a third bit number obtained by adding the first bit number and the second bit number to a least significant bit of a convolution arithmetic operation result output from the convolution arithmetic unit by inputting the target data and the weight after being reduced by the bit reduction unit to the convolution arithmetic unit.

Growing community of inventors

Tokyo, Japan

Daichi Murata