The invention discloses a multi-
modal fusion intelligent video course creation
system, which relates to the technical field of
virtual reality digital
technology system playing and comprises a multi-
modal data input module, a multi-
modal data processing module and a multi-modal course output module. A cross-modal
knowledge graph is constructed by collecting data such as an
industrial equipment operation video, a maintenance manual, a three-dimensional model, tactile feedback and galvanic
skin response, multi-
modal data semantic equivalent conversion is realized, and a new industrial scene is quickly adapted based on a meta-learning
algorithm. The output course comprises videos and texts which are aligned in
time sequence and a three-dimensional model which is subjected to
spatial registration, and a tactile
feedback effect is fused and is optimally presented according to a galvanic
skin reaction. According to the
system, the problems of insufficient practical operation experience and high scene
adaptation cost of traditional training are solved, the training accuracy and efficiency are improved, courses can be applied to
video monitoring,
virtual reality equipment, intelligent consumption equipment and the like, and the
system is suitable for industrial training,
medical teaching and other scenes.