We propose a novel and general framework called the multithreading cascade of Speeded Up Robust Features\n(McSURF), which is capable of processing multiple classifications simultaneously and accurately. The proposed\nframework adopts SURF features, but the framework is a multi-class and simultaneous cascade, i.e., a multithreading\ncascade. McSURF is implemented by configuring an area under the receiver operating characteristic (ROC) curve\n(AUC) of the weak SURF classifier for each data category into a real-value lookup list. These non-interfering lists are\nbuilt into thread channels to train the boosting cascade for each data category. This boosting cascade-based\napproach can be trained to fit complex distributions and can simultaneously and robustly process multi-class events.\nThe proposed method takes facial expression recognition as a test case and validates its use on three popular and\nrepresentative public databases: the Extended Cohn-Kanade, MMI Facial Expression Database, and Annotated Facial\nLandmarks in the Wild database. Overall results show that this framework outperforms other state-of-the-art methods.
Loading....