On July 7, 2021, the final Competition of Best Practices of Applied Algorithms (BPAA) and algorithm summit were held in Zizhu Science and Technology Park, Minhang District, Shanghai. On July 6, the algorithm project, which won the top five of Public, Financial, Medical, Industrial and Commercial tracks in the finals of BPAA algorithm model finals for the full field algorithm award, and BPAA committee presented the trophy to the winning teams. Representatives who attended the event, were Zhang Hongtao, Chief Engineer of Shanghai Municipal Commission of Economy and Information Technology, Guan Xiaojun, member of the Standing Committee and Deputy Head of Minhang District Committee of Shanghai, Du Ping, former Executive Deputy Director of state information center, Li Li, director of Minhang District Science and Technology Commission of Shanghai, and Chen Hongming, Deputy Head of Minhang District Science and Technology Commission of Shanghai.

Photo: Scene of the final competition of BPAA

Sharing frontier knowledge in the arithmetic summit, and aiming to research on the future development trend of the algorithm industry, representatives of this BPAA were, Li Jie, academician of the Japanese Academy of engineering, Chair Professor of Shanghai Jiaotong University and chief scientist of the block chain of Shanghai Artificial Intelligence Research Institute; Yan Junjie, Vice President of Shangtang Technology, Vice President of the Research Institute and head of General Intelligence; Tu Weiwei, Vice President and chief scientist of 4Paradigm; Wu Wei, General Manager of Shanghai Human Intelligence Industry Investment Fund and General Manager of Shanghai Lingang SC-inovation Investment Management Co., Ltd.; Wang Bin, Managing Director of EqualOcean; Song Haitao, General Manager and Executive Director of the Shanghai Institute of Artificial Intelligence; Xie Yongkang, Director Architect of Baidu Intelligent Cloud; Yuan Yue, President of Dataway and other scholars.

Photo: Scene of the BPAA Forum

Zhang Hongtao, Chief Engineer of Shanghai Economic and Information Technology Commission, made a statement in his speech that artificial intelligence is one of the three priority leading industries in Shanghai, and has made great progress in recent years. Since this year, the Shanghai municipal Party committee and government have actively promoted the digital transformation, provided broad application scenarios for the development of artificial intelligence, and accelerated the realization of "economic transformation, lifestyle transformation, governance transformation", so as to provide a new place for the majority of artificial intelligence enterprises and experts. Computing power, computing volume (data) and algorithm are the three basic elements of the development of artificial intelligence. In terms of computing power, Shanghai is actively laying out on the intelligent chip industry and computing power platform, and is building a first-class super large-scale computing power platform in Asia. In terms of data, Shanghai is in the forefront of the country, opening up nearly one billion pieces of public data, forming an all-round and multi-level open trading system for public data and market-oriented data. The state's pilot in data is also being promoted in Shanghai. In terms of algorithms, Shanghai is working out an action plan for algorithm innovation. The first BPAA is a useful practice in promoting the industrialization of algorithms in Shanghai, and is also one of the highlights of the 2021 World Artificial Intelligence Conference. The World Artificial Intelligence Conference will officially open on July 8. The BPAA algorithm model finals and algorithm summit are the first to release relevant achievements accordingly.Gathering advanced resources in the global algorithm field to build a professional, and facilitating international algorithm practice, BPAA Create a professional, open and international platform for algorithm practice, display and exchange, and build a brand new algorithm industry ecological model. In the future, Shanghai will actively encourage and promote the high-quality algorithm projects that stand out in BPAA, and land them in Shanghai, including Minhang District, so as to make good projects and achievements take root in Shanghai.

Photo: Zhang Hongtao, Chief Engineer of Shanghai Economic and Information Technology Commission, delivered a speech

The first BPAA was strongly supported by local government agencies during the seminar. As an important area for the construction of Shanghai Science and technology innovation center and an important support for integrating into the national strategy of Yangtze River Delta integration, Minhang District has attracted a large number of entrepreneurial projects to land for a long time relying on its own advantages such as location conditions, cultural heritage, industrial characteristics and science and technology innovation resources. At the meanwhile, it has formed a strong atmosphere of the algorithm industry, but also to create a complete entrepreneurial service system.

In the speech, Guan Xiaojun, member of the Standing Committee of Minhang District Party committee and Deputy Chief of Shanghai District, said that Minhang District has been leading to deepen the construction of science and technology innovation center in Southern Shanghai, focusing on key areas such as the Big Zero Bay, striving to improve the ability of Science and technology innovation policy source and innovation achievement transformation. At the meanwhile it can strive to build a new benchmark of global innovation and entrepreneurship demonstration, a new highland of national science and technology achievement transformation, a new magnetic field of collaborative innovation in the Yangtze River Delta, and a new platform for scientific and technological innovation in the Yangtze River Delta Shanghai Science and technology innovation center is a new engine of construction. Facing the new opportunities and challenges in the new development stage, Minhang District must make full use of digital mode, explore new ways of social governance in mega cities, and comprehensively promote the urban digital transformation, which has become an important strategy for Shanghai to actively serve the new development pattern, and also an important driving force for the transformation and development of Minhang District. Aiming to speed up the promotion of digital industrialization and industrial digitization, enlarge the radiation and driving role of digital economy, Minhang District helps the development of five oriented economy, fully develops Minhang District's industrial agglomeration advantages, builds a world-class artificial intelligence digital industrial cluster, and a platform for start-up enterprises, and carries out in-depth cooperation with universities and scientific research institutes in the region, so as to shape basic research, business research Industrial two-way link fast lane. BPAA will help the rapid development of digital economy in Minhang District.

Photo: Guan Xiaojun, member of the Standing Committee of Minhang District CPC Committee and Deputy District chief of Minhang District, Shanghai, delivered a speech

Du Ping, former Executive Deputy Director of the National Information Center, stressed in his speech that they should look at algorithm problems, algorithm revolution and algorithm industrialization from the perspective of artificial intelligence and the development process from digital economy to digital intelligence economy. Algorithm existed in ancient time. Benefited by the development of computer science in modern times, algorithm has burst out with strong vitality and huge market potential. As a kind of logical thinking about the objective laws of everything in the world, including human beings. At the same time, algorithm reveals the relationship and causality between the complex big data with the help of computer, seeking solutions to problems. As an indispensable part of artificial intelligence, algorithm is not only a simple tool, but also enables the industrialization of artificial intelligence. The application demonstration of algorithms focused in BPAA is particularly important. Only the technology R & D and scientific and technological innovation that meet a wide range of production needs, consumption needs, social governance needs, government management needs, can have more potential, more value and more market.

Photo: Du Ping, former executive Deputy Director of the State information center, delivered a speech

In the competition performance of BPAA finals, the project of "data driven power battery SOH estimation and fault diagnosis" presented by Shanghai Transportation University in the industrial algorithm track, won the first BPAA award. In recent years, the problem of battery full climate state estimation and spontaneous combustion accidents have become important obstacles to the development of new energy vehicle industry. To solve this problem, the team proposed the concept of data-driven "virtual battery", which captures the dynamic and static information of battery operation through time series data and deep learning algorithm, and achieves the estimation error of health state (SOH) of 1.45% (the highest international accuracy). A fault diagnosis model based on interpretable electrochemical parameters is proposed, and a machine learning algorithm is constructed to train the distribution of electrochemical parameters, so as to realize accurate diagnosis of battery operation fault and early warning of thermal runaway 48 hours in advance. And actively carried out the exploration and practice of industrialization.

Photo: Awarded the BPAA Grand Prize to the winning team

At the same time, other teams participating in the BPAA finals also performed well. As a public algorithm track, the project of "Segmentation Transformers" presented by Fudan University, the tension research group has put forward a novel a new neural network paradigm SEgmentation TRansformer (SETR), based on self attention mechanism sequence prediction for semantic segmentation task. It has abandoned full convolution neural network in a trans-formative way. The algorithm achieves the best performance on several international authoritative semantic segmentation datasets, including MIT Scene Parsing Benchmark ADE20k and Pascal Context. This work subverts the dominant position of the traditional full convolution neural network in the dense prediction task of computer vision for many years, and provides a new solution and new research direction for computer vision and core tasks.

Photo: Scene of BPAA

In the financial algorithm track, the project of "intelligent claim verification solution based on artificial intelligence algorithm" presented by Wanda Information Co., Ltd. With a focus on the problem of slow payment of commercial insurance, it realizes fast claim settlement through intelligent data docking with medical institutions and medical insurance bureau, and provides intelligent claim material processing capability based on machine vision for institutions that have not yet done data docking. Up to now, it focus on the research and development of intelligent recognition and structured model of paper medical bills with high accuracy, which improves the efficiency of claim processing by more than 80%, and has served many large insurance companies.

Photo: Scene of BPAA

The project of "UKnow hemorrhagic stroke AI solution / UGuard for ischemic stroke AI solution" presented by Union strong (Beijing) Technology Co., Ltd, it indicated that, cerebrovascular disease has become a major killer of human health in the advent of aging society. Taking cerebral aneurysms as an example, the incidence rate of Chinese cerebral artery can reach 5 to 7%. Meanwhile, the treatment cost of cerebral aneurysm rupture is huge and the mortality rate is high. Therefore, it is particularly important for early diagnosis, comprehensive evaluation and precise treatment of cerebral aneurysms. In order to solve the difficulties and pain points in the screening, diagnosis and treatment of cerebrovascular diseases that seriously endanger human life and health, such as the timeliness of screening and prevention, the consistency of diagnostic criteria, the diversity of preoperative consumables planning, the difficulty of establishing intervention access, the accuracy of prognosis judgment, etc. The project adopts advanced technologies such as deep learning, image processing, simulation and so on, and eventually combined with a series of clinical application research and multi center application research, the intelligent diagnosis and treatment platform for stroke is finally formed.

Photo: Scene of BPAA

The project of "industry leading artificial intelligence document translation algorithm" presented by Sichuan Yixun Information Technology Co., Ltd. Which is a technology company focusing on cross language big data and artificial intelligence technology. It is committed to utilize the international cutting-edge natural language processing and semantic processing technology to provide global users with scenario application solutions combining AI with segmentation industry. The founding team of Yixun Technology is composed of language experts who have been engaged in translation industry for 18 years, such as AI doctors from Hong Kong and algorithm scientists who have returned from the United States. The R&D team is composed of software engineers from Microsoft, Huawei, Symantec, ThoughtWorks, etc. Through the construction of "global language big data center", Yixun Technology has more than 10 billion bilingual translation corpus databases (second only to Google), especially in the level of manual translation database with Chinese as the core language, which has obvious advantages in the industry.

Photo: Scene of BPAA

On the morning of July 7, the top 50 teams of BPAA, visited Hongqiao Business District, Shanghai Maqiao Artificial Intelligence Innovation Experimental Zone, and innovation and entrepreneurship clusters in Big Zero Bay. As a science and innovation center in the south of Shanghai, a national achievement transfer and transformation demonstration area, and a national intellectual property demonstration area, Minhang District is rich in science and innovation resources and scientific and technological achievements. Minhang District has always attached great importance to the promotion of science and technology support and industrial competitiveness. Its science and innovation policy gives full play to the innovation resources and high-level innovation ability of the project, and works with all walks of life to provide support for technological innovation and industrial innovation. The relevant science and innovation policy subjects in Minhang District focus on four aspects: First, focus on the enterprise growth chain; Second, focus on enterprise innovation ability; Third, focus on the transformation of scientific and technological achievements; Fourth, focus on optimizing the ecology of scientific innovation. In the future, Minhang District will be built into the core of entrepreneurship cluster in the field of algorithm, the hub of resource integration and sharing, and a new platform for innovation and entrepreneurship teams to achieve mutual benefit and win-win with local governments. Excellent algorithm teams are welcome to join in Minhang District for innovation.

Photo: Chen Hongming, Deputy Director of science and Technology Commission of Minhang District, Shanghai, introduced relevant science and innovation policies of Minhang District

Li Jie, academician of the Japanese Academy of engineering, lecturer professor of Shanghai Transportation University and chief scientist of block chain of Shanghai Institute of Artificial intelligence, delivered a speech on the topic of Algorithm Innovation Driving AI Enabled Smart City Construction at the algorithm summit. He propose to build a multi-dimensional, multi-scale, multi-source, cross media, virtual and real integration dynamic modeling and simulation system and platform, so as to support the city's digital virtualization of all elements, real-time visualization of all States, collaborative and intelligent management and decision-making of the city.

Yan Junjie, vice president of Shangtang Technology, Vice President of Research Institute and head of general intelligence, addressed in speech entitled as Moore's Law of General Artificial Intelligence, that collaborative innovation should be based on algorithm, computing power and system to promote the exponential improvement of the performance of general artificial intelligence model.

In the speech entitled Data Driven Decision Optimization: Environmental Learning and Reinforcement Learning, Tu Weiwei, Vice President and Chief scientist of 4Paradigm, advocated that building a hierarchical decision-making system of human-computer cooperation will achieve the future of AI algorithm blessing.

Yuan Yue, chairman of Dataway, delivered a speech entitled the Growth and Cultivation of Algorithms. Through the development and formation of algorithms and algorithm sets, he vividly demonstrated the whole process of algorithm products from conceptual architecture to decision-making brain core. He also stressed that the key to promoting intelligent and intelligent industry is to jointly promote the development of algorithm production factors and make the algorithm industrialization truly achieve the realization of energy in Shanghai, through the concerted efforts of all parties.

Photo: Scene of the BPAA forum

Photo: Yuan Yue introduced BPAA project to Zhang Hongtao and Guan Xiaojun

Photo: Visit to the Key Innovation Industrial Park in Minhang District