|
Abstract
|
The paper explores the application of higher order directional derivatives into Local Binary Patterns (LBP) for facial expression recognition. It introduces a novel method that extends the standard LBP, typically utilizing only first order directional derivatives, to include second order derivatives. The study highlights the importance of the number of directions incorporated in this new approach. As a result, compelling findings are achieved, emphasizing the significance of this research area.
|