A major Japan company has developed a new technology for utilizing generative AI.
It is a technology in which AI recognizes the content of a video and expresses it in writing, and it is likely that the range of use of AI will expand, such as analyzing the cause of traffic accidents from the video of the drive recorder.
Generative AI is said to be more technically difficult to recognize videos with a large amount of information, unlike words and images.
Against this backdrop, NEC has developed a new technology that allows AI to recognize the content of a video on its own and express it in text.
It is characterized by the cooperation of more than 100 AIs to recognize videos, so to speak, and the AI, which is good at recognizing each person and object, first transcribes them as fragmentary words. It is a mechanism in which the generation AI compiles the recognition results and arranges them into sentences.
For example, this technology is expected to be used to analyze the causes of traffic accidents from dashcam videos and create reports, or to create work records from videos of construction sites.
The technology for recognizing video by generative AI is also being developed by major IT companies in the United States, but it is rare to combine a large number of AIs, so it will be interesting to see if the presence of major companies in the Japan will increase in competition in new fields.