Facial Expression Analysis with Attention Mechanism

Facial expressions are configurations of different muscle movements in the face. The local characters of muscle movements play an important role in distinguishing facial expressions by machines. In this webinar, the presenter will explore the local characters local characters of muscle movements by introducing the attention mechanism into two frameworks: 1) Propose a supervised convolutional neural network (CNN) with attention mechanism to recognize the facial expression with partially occluded faces. Through the attention module, the weights for different facial regions were understood along with the perceived occluded regions of the face and focused on the most discriminative unoccluded regions. 2) Propose a self-supervised facial action representation learning framework, where the attention mechanism is embedded in the encoder. Through the attention module, the result was to be able to discover the discriminative facial regions in an unsupervised manner and the ability to achieve self-supervised representation that improves the performance of facial action unit detection.