An Efficient Bayesian Approach to Exploit the Context of Object-Action Interaction for Object Recognition
This research features object recognition that exploits the context of object-action interaction to enhance the recognition performance. Since objects have specific usages, and human actions corresponding to these usages can be associated with these objects, human actions can provide effective infor...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2016-06-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | http://www.mdpi.com/1424-8220/16/7/981 |
| _version_ | 1828117816565301248 |
|---|---|
| author | Sungbaek Yoon Hyunjin Park Juneho Yi |
| author_facet | Sungbaek Yoon Hyunjin Park Juneho Yi |
| author_sort | Sungbaek Yoon |
| collection | DOAJ |
| description | This research features object recognition that exploits the context of object-action interaction to enhance the recognition performance. Since objects have specific usages, and human actions corresponding to these usages can be associated with these objects, human actions can provide effective information for object recognition. When objects from different categories have similar appearances, the human action associated with each object can be very effective in resolving ambiguities related to recognizing these objects. We propose an efficient method that integrates human interaction with objects into a form of object recognition. We represent human actions by concatenating poselet vectors computed from key frames and learn the probabilities of objects and actions using random forest and multi-class AdaBoost algorithms. Our experimental results show that poselet representation of human actions is quite effective in integrating human action information into object recognition. |
| first_indexed | 2024-04-11T13:21:48Z |
| format | Article |
| id | doaj.art-bc8b0c470e704faea77df297eab15bcb |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-04-11T13:21:48Z |
| publishDate | 2016-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-bc8b0c470e704faea77df297eab15bcb2022-12-22T04:22:11ZengMDPI AGSensors1424-82202016-06-0116798110.3390/s16070981s16070981An Efficient Bayesian Approach to Exploit the Context of Object-Action Interaction for Object RecognitionSungbaek Yoon0Hyunjin Park1Juneho Yi2School of Electronic and Electrical Engineering, Sungkyunkwan University, Suwon 16419, KoreaSchool of Electronic and Electrical Engineering, Sungkyunkwan University, Suwon 16419, KoreaSchool of Electronic and Electrical Engineering, Sungkyunkwan University, Suwon 16419, KoreaThis research features object recognition that exploits the context of object-action interaction to enhance the recognition performance. Since objects have specific usages, and human actions corresponding to these usages can be associated with these objects, human actions can provide effective information for object recognition. When objects from different categories have similar appearances, the human action associated with each object can be very effective in resolving ambiguities related to recognizing these objects. We propose an efficient method that integrates human interaction with objects into a form of object recognition. We represent human actions by concatenating poselet vectors computed from key frames and learn the probabilities of objects and actions using random forest and multi-class AdaBoost algorithms. Our experimental results show that poselet representation of human actions is quite effective in integrating human action information into object recognition.http://www.mdpi.com/1424-8220/16/7/981object recognitionobject-action contextobject-human interaction |
| spellingShingle | Sungbaek Yoon Hyunjin Park Juneho Yi An Efficient Bayesian Approach to Exploit the Context of Object-Action Interaction for Object Recognition Sensors object recognition object-action context object-human interaction |
| title | An Efficient Bayesian Approach to Exploit the Context of Object-Action Interaction for Object Recognition |
| title_full | An Efficient Bayesian Approach to Exploit the Context of Object-Action Interaction for Object Recognition |
| title_fullStr | An Efficient Bayesian Approach to Exploit the Context of Object-Action Interaction for Object Recognition |
| title_full_unstemmed | An Efficient Bayesian Approach to Exploit the Context of Object-Action Interaction for Object Recognition |
| title_short | An Efficient Bayesian Approach to Exploit the Context of Object-Action Interaction for Object Recognition |
| title_sort | efficient bayesian approach to exploit the context of object action interaction for object recognition |
| topic | object recognition object-action context object-human interaction |
| url | http://www.mdpi.com/1424-8220/16/7/981 |
| work_keys_str_mv | AT sungbaekyoon anefficientbayesianapproachtoexploitthecontextofobjectactioninteractionforobjectrecognition AT hyunjinpark anefficientbayesianapproachtoexploitthecontextofobjectactioninteractionforobjectrecognition AT junehoyi anefficientbayesianapproachtoexploitthecontextofobjectactioninteractionforobjectrecognition AT sungbaekyoon efficientbayesianapproachtoexploitthecontextofobjectactioninteractionforobjectrecognition AT hyunjinpark efficientbayesianapproachtoexploitthecontextofobjectactioninteractionforobjectrecognition AT junehoyi efficientbayesianapproachtoexploitthecontextofobjectactioninteractionforobjectrecognition |