Integration of Multiple-Modality Sensor Data and Artificial Intelligence

Multimodal perception is not only an important ability of human intelligence, but also one of the essential differences between human intelligence and artificial intelligence. With the rapid development of sensor technology, artificial intelligence technology and the Internet, various modalities of sensor data are emerging rapidly, for example, vision sensors and voice sensors are widely used in target detection and speech interaction. Integrating multiple-modality sensor data could obtain more comprehensive and accurate information, and also enhance the reliability and fault tolerance of the system. Clarifying the principles and mechanisms of how human brain integrates multi-sensory information can provide a theoretical basis for the scientific development of artificial intelligence. The introduction of multi-modality integration modules in artificial intelligence can simulate human perception largely, thus artificial intelligence will be infinitely similar to human intelligence.