| Paper | Conference | Remarks |
|---|---|---|
| FaceWarehouse: A 3D Facial Expression Database for Visual Computing | IEEE TVCG 2013 | 1. RGBD data from 150 individuals aged 7-80 from various ethnic backgrounds. 2. Neutral and 19 other expressions are captured for each person. 3. For every RGBD raw data record, a set of facial features points are located. 4. Match feature points on color images to depth image mesh. 5. Construct a set of individual-specific expression blendshapes for each person. 6. A bilinear face model with attributes: identity and expression. 7. Advantages over other databases: much richer matching collection of expressions. 8. Applications: facial image manipulation; face component transfer; real-time performance-based facial image animation; facial animation retargeting from video to image |
| Using Kinect for real-time emotion recognition via facial expressions | FIETT 2015 | 1. A real-time emotion recognition approach based on both 2D and 3D facial expression features captured by Kinect sensors. 2. To capture the deformation of 3D mesh during facial expression, the features of animation units and feature point positions are combined. 3. A fusion algorithm based on improved emotional profiles (IEPs) and maximum confidence is proposed to recognize emotions. 4. Show superior performance of our method. |