Detecting and Grouping Identical Objects for Region Proposal and Classification

TL;DR

This paper introduces a method that detects and groups identical objects in scenes to improve region proposals and classification accuracy, enabling hierarchical classification and reducing computational load.

Contribution

It presents an unsupervised multi-instance object discovery approach integrated with CNN classification, enhancing efficiency and enabling sub-classification of object types.

Findings

Fewer regions evaluated compared to traditional methods

Improved classification accuracy through joint probability of object instances

Enables hierarchical classification by splitting classes into sub-classes

Abstract

Often multiple instances of an object occur in the same scene, for example in a warehouse. Unsupervised multi-instance object discovery algorithms are able to detect and identify such objects. We use such an algorithm to provide object proposals to a convolutional neural network (CNN) based classifier. This results in fewer regions to evaluate, compared to traditional region proposal algorithms. Additionally, it enables using the joint probability of multiple instances of an object, resulting in improved classification accuracy. The proposed technique can also split a single class into multiple sub-classes corresponding to the different object types, enabling hierarchical classification.