Method and Apparatus for Recognizing Objects in a Weakly Supervised Learning Manner

摘要

A method and apparatus for object recognition in a weakly supervised learning manner are provided. A method of recognizing a weakly supervised learning method performed by an object recognition apparatus includes a step of extracting a plurality of feature maps from a given training object image with a classification result of an object of interest, Extracting a plurality of feature maps from the plurality of feature maps, generating an activation map for each of the objects of interest by accumulating the plurality of feature maps, synthesizing the activation values included in the activation map of the object of interest, Determining an error by comparing a classification result determined using the representative value of the object of interest with the given classification result, and back-propagating the error to generate a convolutional neural network Based object recognition model.