Giants have laid out one after another, and the market has also attracted more and more talents to start their businesses. Computer vision is becoming one of the hottest areas of artificial intelligence.
This report will analyze and study the key nodes, market status, and application scenarios of computer vision technology development.
I. Analysis of technological development and market status
1. Artificial intelligence is a technological evolution from the ultimate concept to classification and landing
2. The infrastructure on which artificial intelligence relies is already in place, but it is still at an early stage
Artificial intelligence is growing like a baby. Machines are no longer just accomplishing tasks through specific programming, but can master skills through continuous learning. This mainly relies on efficient model algorithms for large amounts of data training, which requires high-performance computing behind them. Ability of hardware and software as support. With the rapid development of the Internet and the continuous advancement of the underlying technologies, the "energy" required for artificial intelligence is constantly improving.
Data volume: Since 2000, the rapid development of the Internet and mobile Internet has resulted in the accumulation of data. According to IDC's forecast, the total amount of global data in 2020 will be 40ZB, of which 70% will be conducted in the form of pictures and videos. Storage, which provides a rich soil for the development of artificial intelligence.
Deep Learning Algorithm: University of Toronto professor Geoffrey Hinton (works on neural network and deep learning research) students in the industry-renowned image recognition competition ImageNet used deep learning algorithms to reduce the recognition error rate by 10%, even exceeding Google's depth. Learning then became famous. In 2015, the Visual Computing Group of Microsoft Research Asia won the championship in this competition, reducing the system error rate to 3.57%, which has exceeded the human eye.