A robust automatic micro-expression recognition system would have broad applications in national safety, police interrogation, and clinical diagnosis. Developing such a system requires high quality databases with sufficient training samples which are currently not available. We reviewed the previously developed micro-expression databases and built an improved one (CASMEII), with higher temporal (200fps) and spatial resolution (about 280x340 pixels on facial area). We elicited participants' facial expressions in a well-controlled laboratory environment, with proper experimental design and illumination. Among nearly 3000 facial movements, 247 micro-expressions were selected for the database with action units (AUs) labeled. For baseline evaluation, LBP-TOP and SVM were employed respectively for feature extraction and classifier with the leave-one-subject-out cross-validation method. The best performance is 63.41% for 5-class classification.

The CASME II database (http://fu.psych.ac.cn/CASME/casme2-en.php) has the following characteristics:

• The samples are spontaneous and dynamic micro-expressions. Baseline (usually neutral) frames are kept before and after each micro-expression, making it possible to evaluate different detection algorithms.
• The recordings have high temporal resolution (200 fps) and relatively higher face resolution at 280x340 pixels.
• Micro-expression labeling is based on FACS investigator's guide and Yan et al. findings (Yan et al., 2013) that is different from the traditional 6 categories on ordinary facial expression.
• The recordings have proper illumination without lighting flickers and with reduced highlight regions of the face.
• Some types of facial expressions are difficult to elicit in laboratory situations, thus the samples in different categories distributed unequally, e.g., there are 60 disgust samples but only 7 sadness samples. In CASME II, we provide 5 classes of micro-expressions.