Smile detection from still images using KNN algorithm
| Autor: | Sneha Jose, Sumi. P. Potty, Treesa George |
|---|---|
| Rok vydání: | 2014 |
| Předmět: |
Facial expression
business.industry Computer science ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Pattern recognition k-nearest neighbors algorithm Object-class detection Open source Lazy learning Smile detection Computer vision Artificial intelligence business Classifier (UML) Blossom algorithm |
| Zdroj: | 2014 International Conference on Control, Instrumentation, Communication and Computational Technologies (ICCICCT). |
| DOI: | 10.1109/iccicct.2014.6993006 |
| Popis: | Reliable detection and recognition of facial expression from still images in the unconstrained real world situations has many potential applications. Smile detection can be used in many applications include modeling systems for psychological studies on human emotional responses, expression recognition technologies, extending image search capabilities etc. This paper proposes an experimental study of smile detection in embedded environment using Raspberry Pi board, by extracting mouth and eye pair from images using Haar-cascade classifier and train these images using KNN matching algorithm. The relatively simple K- Nearest Neighbor is used because of its lazy learning efficiency. OpenCV- 2.3.1(Open Source Computer Vision) library is used as the imaging library. The experiments explored that the proposed approach has an accuracy of 66.6%. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|