Huawei Computer Vision Plan: Full-Stack, All-Scenario Industry AI


Owing to the advancement of algorithms combined with deep learning, Industry AI will significantly help businesses or public institutions increase the work efficiency. In medical research AI is also expanding its function. Take the retinal image in the fundus as an example. Doctors can assess the health of body organs simply by analyzing blood flow in the retinal system, which are the naked eye's only blood vessels visible. This method is supercharging more reliable and less time-consuming medical diagnosis thanks to new AI technologies. This sentiment is shared by Professor Li Tao, a specialist at Nankai University in intelligent fundus retinal images computing, who noted how AI-assisted diagnosis and treatment would help save more lives. Artificial intelligence is changing the world as we know it, from human intelligence's simple learning measures to pioneering medical advances.
Computing and algorithms: A powerful duo
Strong computing, sophisticated algorithms, and mass data are the cornerstone for efficient, wide-ranging AI industry. Also crucial to boosting AI application across industries are innovations, debates, and data sharing between companies and academia.
Professor Li, who collaborates with Huawei to create smart computing designed for fundus retinal images, is demonstrating an example of this crossover. At HDC 2020, Professor Li shared how AI creativity is supercharged by the Huawei Atlas 200 DK Developer Kit running on Ascend Ai processors. By delivering ultimate input power at low power consumption, Atlas 200 DK effectively funds end-to-end scientific science. Luckily for Professor Li and the medical community, this feat is not confined to Atlas 200 alone, with Huawei offering a comprehensive range of Atlas products that unleash the synergy between algorithms and data to maximize efficiency.
Huawei Computer Vision Plan
Huawei released the Computer Vision Plan at the Huawei Developer Conference 2020 (Cloud), and invited global AI experts to take part in the research. To support this plan, the Atlas AI computing platform powered by Huawei Ascend AI processors will deliver powerful computing. The research findings will be incorporated in Huawei's MindSpore, the industry-opened all-scenario AI computing platform, allowing global AI developers to continually evolve, break barriers and create pervasive intelligence.
An important aspect of Huawei's AI Strategy is innovation in fundamental research. Huawei has developed basic skills in areas such as computer vision, natural language processing, decision making and inference. These capabilities are critical to the development of efficient, green, stable, reliable and autonomous machine learning.

Tian Qi, Computer Vision Chief Scientist, Huawei's Noah's Ark Laboratory and IEEE Fellow, shared the latest developments in computer vision science. "Huawei has invested heavily in computer vision basic research focussing on data, information, and models. Huawei has written over 80 articles in leading AI conferences and journals such as CVPR, ICCV, NeurIPS, And to the ICLR. Huawei's made other revolutionary achievements. Such work findings in the form of academic papers and source code are accessible to industry. We are inviting global AI developers to research, develop, and deploy AI based on existing research results from Huawei.
The Computer Vision Plan incorporates six sub-plans. They are:
Ø Data Iceberg Plan: Use a small amount of annotated data to unleash the potential of massive unannotated data and support model training in small sample scenarios.
Ø Data Magic Cube Plan: Use multi-modal quantification, alignment, and fusion strategies to enhance the learning capability of models in real-world scenarios.
Ø Model High-Touching Plan: Build large models on the cloud to explore the performance limits of various vision tasks.
Ø Model Slimming Plan: Build efficient computing models on the device side to help various chips complete complex inference.
Ø Generic Vision Plan: Define vision pre-training tasks to build generalized vision models.
Ø V-R Integration Plan: Direct computer vision to real artificial intelligence via virtual-real integration.
We live in a world that demands ubiquitous cloud and widespread AI and for any technical player a prosperous Al ecosystem is a must.

Post a Comment

0 Comments