Abstract | ||
---|---|---|
This paper presents a multi-pose face registration method for social robot application. A social robot means that it is an autonomous robot to interact and communicate with humans. The first thing in communicating with people is to recognize who they are. To do this, the social robot should basically have a face recognition function. Although many face recognition algorithms have been developed in the past, the development of algorithms that are robust to real-time pose changes is underway. In this paper, we try to a multi-pose face registration method for pose invariant face recognition. To measure the robustness of the proposed method, comparisons were made between the registering of front face only versus the registering multiple pose faces based on their respective recognition similarity values. As a result, it was confirmed that the confidence value of similarity always keeps a high value when the proposed method is used compared to when not using it, despite the fact that the face was entered in various poses. |
Year | Venue | Field |
---|---|---|
2018 | ICSR | Social robot,Facial recognition system,Computer vision,Face registration,Computer science,Robustness (computer science),Invariant (mathematics),Artificial intelligence,Autonomous robot |
DocType | Citations | PageRank |
Conference | 0 | 0.34 |
References | Authors | |
17 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Ho-Sub Yoon | 1 | 11 | 3.88 |
Jaeyoon Jang | 2 | 0 | 0.34 |
Jaehong Kim | 3 | 383 | 41.59 |