Potential emotion recognition method based on knowledge guidance and generative adversarial learning

• Main Authors: MA JUNHUI, NIU JIN, SONG JIANQIAO, JIA HAIRONG, WANG FEI, ZHAO CHENCHEN, XIANG CHENZHUO, XIANG HUSHENG, WANG FENG, SHI ZEZHOU
• Format: Patent
• Language: Chinese; English
• Published: 11.08.2020
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS

The invention discloses a potential emotion recognition method based on knowledge guidance and generative adversarial learning. The method mainly comprises the following steps of: 1, watching an induction video of a corresponding emotion for a subject so as to enable the subject to generate a corresponding potential emotion, using a Neuroscan Synamps2 electroencephalogram amplifier to collect EEGelectroencephalogram signals, and using a camera to collect facial expression video data to facilitate subsequent processing to obtain BVP pulse signals; 2, building a generative adversarial neural network model, and building a generator and a discriminator; and 3, performing discrimination by using the trained discriminator, and performing D-S decision fusion on discrimination results output by the two generative adversarial network discriminators to obtain a final result. The potential emotion recognition method based on knowledge guidance and generative adversarial learning has the advantages of being simple, effic