The name of the base system (platform): | Artificial intelligence (AI, Artificial intelligence, AI) |
Developers: | Huawei |
Date of the premiere of the system: | 2017/09 |
Technology: | BI, Cloud Computing, EDMS - Systems of stream recognition |
Huawei Enterprise Intelligence (EI) is the intellectual platform as a part of which — three types of cloud services and the heterogeneous computing platform.[1]
Basic services of the platform:
- machine learning
- deep learning
- analysis of graphs
- training of AI
- justification of proposed solutions
- indexation
The AI general services include API of recognition of a voice, visual recognition and a natural language understanding.
Scenarios of application of EI:
- processing of video in real time,
- identification of simple objects,
- analysis of documents and images.
2017: Presentation
At the beginning of September, 2017 Huawei, during the annual Huawei Connect conference in Shanghai, submitted the Huawei Enterprise Intelligence (EI) platform using which the company intends to provide the main services of deep training, the analysis of graphs, artificial intelligence, machine learning, justification of proposed solutions and indexation.[2]
The platform supports API of general purpose, including recognition of a voice and a natural language. Accuracy of optical character recognition, by estimates of Huawei, as of September, 2017 is 97.37%.
According to plans of the company, the Enterprise Intelligence platform will be applied when implementing initiatives of "the smart cities". In the systems of regulation of traffic she is capable to process to 100 billion frames per second.
In Huawei gave the Traffic Brain project implemented in Shenzhen as an example. The small fishing village turned into the huge city with the population in 22 million people, and fight against traffic jams is very relevant here. The city authorities are going to use Huawei network with ultrahigh capacity for data analysis about traffic. The system expected registration of 700 million records monthly will process every day 10 million images. It is expected that technologies will allow to reveal vehicles which should not be on the road.