Visual localization in images using weakly supervised neural network

Abstract

A system and method for visual anomaly localization in a test image includes generating, in plurality of scaled iterations, attention maps for a test image using a trained classifier network, using image-level. A current attention map is generated using an inversion of the classifier network on a condition that a forward pass of the test image in the classifier network detects a first class. One or more attention regions of the current attention map may be extracted and resized as a sub-image. For each scaled iteration, extraction of one or more regions of a current attention map is performedon a condition that the current attention map is significantly different than the preceding attention map. Visual localization of a region for the class in the test image is based on one or more of the attention maps

Description

technical field [0001] The present invention relates to artificial intelligence. More specifically, the present invention relates to the application of artificial intelligence to visual recognition systems. Background technique [0002] Visual recognition systems can apply machine learning-based methods, such as convolutional neural networks, which can involve training the system to recognize features or objects of interest in images based on learned classifications. Classification can include tangible attributes, such as identifying the presence of a particular animate or inanimate object. For example, a system can be trained to learn one or more classifications (eg, flowers, dogs, chairs), and once trained, it analyzes a series of test images to identify which images include objects of the trained classifications. [0003] Visual localization with machine learning assistance can be applied to detect anomalies within images or identify unusual objects potentially causing ...

AI Technical Summary

Problems solved by technology

[0004] Current approaches to machine learning-based visual localization have practical limitations, including requiring intensive manual pixel-wise or bounding-box labeling of training images

For example, labeling can include drawing bounding boxes around unusual objects that appear in the image, which is time-consuming and non-scalable

Images