Human-computer collaborative object recognition for intelligent support

Authors:
Kazuaki Kondo;Hideyuki Nishitani;Yuichi Nakamura
Affiliations:
Academic Center for Computing and Media Studies, Kyoto University, Kyoto, Japan;Academic Center for Computing and Media Studies, Kyoto University, Kyoto, Japan;Academic Center for Computing and Media Studies, Kyoto University, Kyoto, Japan
Venue:
PCM'10 Proceedings of the Advances in multimedia information processing, and 11th Pacific Rim conference on Multimedia: Part II
Year:
2010

Citing 3
Cited 0

Text Classification from Labeled and Unlabeled Documents using EM

Machine Learning - Special issue on information retrieval
Distinctive Image Features from Scale-Invariant Keypoints

International Journal of Computer Vision
Semi-automatic photo annotation strategies using event based clustering and clothing based person recognition

Interacting with Computers

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper introduces a novel framework for collaborative object recognition, which expands the applicability and improves the accuracy of object recognition. In this framework, a system not only recognizes targets but also detects and evaluates conditions that may make recognition difficult, and tries to resolve the situation by presenting the user with information on how to alter the conditions. The user can see how to make improvements, leading to correct recognition with little effort. The system can provide a useful, easy-to-use tool. In this research, a prototype system for kitchen scenes is designed, which can achieve situation evaluation and human-computer collaboration to improve recognition. We verified the framework by observing improvements in recognition accuracy and behavior of users in our experiments.