B	No-pain
E1	Low electrical-pain stimuli
E2	Moderate electrical-pain stimuli
E3	Severe electrical-pain stimuli
H1	Low heat-pain stimuli
H2	Moderate heat-pain stimuli
H3	Severe heat-pain stimuli
AU	Action Unit
Conv	Convolution layers
CNN	Convolutional Neural Network
CNNs*	Sample weighting method used with two simple convolutional Neural Network
ECG	Electrocardiogram
EDA	Electrodermal activity
EMG	Electromyography
FACS	Facial Action Coding System
FAD	Facial Activity Descriptor
MVV2	MobileNetV2
MNV2*	Sample weighting method used with MobileNetV2 and simple convolutional Neural Network
MTL	Multi-task learning
MTL-CNNs-loss-function (softmax, softmax)	The trained Multi-task learning model using softmax loss with two simple convolutional Neural Network
MTL-CNNs-loss-function (softmax, mse)	The trained Multi-task learning model using softmax and mse loss with two simple convolutional Neural Network respectively
MTL-CNNs-loss-function (softmax, sigmoid)	The trained Multi-task learning model using softmax and sigmoid loss with two simple convolutional Neural Network respectively
MTL-CNNs*-loss-function (softmax, softmax)	The trained Multi-task learning model using softmax loss with two simple convolutional Neural Network
MTL-CNNs*-loss-function (softmax, mse)	The trained Multi-task learning model using softmax and mse loss with MobileNetV2 and simple convolutional Neural Network respectively