Abstract:AI development is beautiful to look at, but not easy to put into practice. A smart developer knows how to use tools to improve development efficiency, while a smart platform will stand on the developer’s side and provide intimate services to users.
preface
“The ideal is full, the reality is skinny.” There is no better sentence to describe the state of AI development than this one. In the past two years, artificial intelligence, as a sunrise industry, has attracted capital, enterprises and individuals to flood in, and has a tendency to land on the beach. Especially for geeky developers, who want to ride the wave of AI development and be a trendsetter.
However, while AI development is beautiful to look at, it is not easy to put into practice. Especially in this era of rapid information explosion and technological iteration, data is growing exponentially, models are becoming more and more complex, and the calculation process is always time-consuming and laborious… All these problems are plaguing dream-seeking AI development friends, but also hinder the development of artificial intelligence industry.
A smart developer knows how to use tools to improve development efficiency, while a smart platform will stand on the developer’s side and provide intimate services to users. ModelArts is such an intelligent and intimate platform.
The voice of the developer, Ta knows
As we know, the needs of AI development tools vary greatly among different groups of people, such as xiaobai and Danniu. ModelArts is sweet because it can meet different needs of different users:
For application developers who desperately need AI capabilities but do not have AI capabilities, ModelArts can help them generate AI models and services from zero code; For AI beginners, ModelArts can help them get started quickly; For senior AI developers, ModelArts can help them efficiently obtain and utilize AI computing resources, manage the whole development process, and let them focus on AI development without worries.
ModelArts is a one-stop AI development platform for developers, which was released by Huawei Cloud in October 2018. ModelArts provides data processing, intelligent labeling, out-of-the-box development environment, large-scale distributed training, automatic model generation, rich AI market, and complete AI development capability of end-to-end edge-to-cloud model deployment in all scenarios, helping users to create and deploy AI models quickly and efficiently, and managing the entire development life cycle.
In order to provide AI developers with more intimate services, Huawei Cloud released Models Arts2.0 in September 2019. ModelArts2.0 upgrades the existing AI development mode with the minimalism and automation of the whole process, so that the whole chain of data preparation, algorithm development, model training, model management and model reasoning has a qualitative leap.
Compared with version 1.0, ModelArts2.0 released more than ten new features and services, including intelligent data screening, intelligent data annotation, intelligent data analysis, multiple model automatic search, ModelArts SDK, graph neural network, reinforcement learning, model evaluation/diagnosis, model compression/transformation, difficult cases automatic discovery, online learning, etc. Cover the entire AI model lifecycle.
Among them, intelligent data screening means to automatically filter and screen out the data invalid to the training model by AI. Hybrid intelligent labeling technology can improve the labeling efficiency by at least 5 times; Combined with online learning mode, the automatic discovery function of difficult cases can open up the running state and development state, so that the model can continue to learn and evolve online with the growth and change of data, and truly provide evolvable AI capabilities.
As is known to all, the evaluation and tuning of model performance is an important and difficult task for algorithm developers. ModelArts2.0 provides a rich model evaluation interface and visualization capabilities, and can provide model diagnosis suggestions to AI developers based on the evaluation results to help developers reduce the difficulty of tuning.
Never stop doing what you love
Pan, a student from Southwest University, has been operating an artificial intelligence public account since his university. What you really love is never tiring. He hopes that through his own small help, the public will pay more attention to AI, understand AI, and participate in AI. From the very beginning, I got in touch with ModelArts, a one-stop AI development platform of Huawei cloud, and tried out each product. I shared my own experience and carefully sorted out each document to provide convenience for “latecomers”.
With the attention and encouragement of more people, Pan began to attack the actual combat case articles, bringing more professional content to share for everyone, so that each practical application case let everyone better know and understand Huawei cloud one-stop AI development platform -ModelArts.
Love a thing is not just ye gonglong, it is his sincere pay. Received the invitation of Huawei cloud community and made an online live broadcast for AI developers across the country. “At present, I am constantly recording the classic cases of ModelArts and adjusting my experience. We hope that more developers can experience the charm of ModelArts, huawei’s cloud one-stop AI development platform.” When he got to know more developers, pan became an expert of Huawei cloud sharing thanks to his constant efforts and love.
To play together is better than to fight alone
AI development is a long way off, and everyone wants to improve their skills and grow into a god-level developer. And just like martial artists, to improve their skills more quickly, it is often better to learn from each other than to practice alone. Huawei Cloud Base builds the developer ecological community in ModelArts, providing a safe and open sharing and trading environment for university research institutions, AI application developers, solution integrators, enterprises and individual developers, etc., effectively connecting all participants in the AI development ecological chain, and accelerating the development and implementation of AI products.
Take Jinyu Medical as an example. In cooperation with Huawei Cloud AI team, jinyu Medical pathologists trained an accurate and efficient AI-assisted cervical cancer screening model based on pathological morphology and deep learning technology. The accuracy rate of negative film interpretation was higher than 99% and the detection rate of positive lesions was higher than 99.9% on the basis of the rejection rate higher than 60%. With the aid of AI, the pathological interpretation of each case only takes 36 seconds. ModelArts provides a solid foundation for the training and application of AI models from the aspects of annotation, training model, image judgment and inference.
Currently ModelArts developer ecological community has 100000 +, and with more than 50 colleges and universities to cooperate, and gradually applied in medical treatment, the Internet, the wisdom city, building, garden and so on ten big industry 600 + project, help golden field medical, YunLu technology, visionbank of enterprise and the development and application developers to AI.
Buildings constitute the basic unit of a city and are important places for people to engage in production and life. The quality of construction projects is closely related to the safety of people’s life and property. Yunlu technology, a top service provider in the field of engineering monitoring, has joined hands with Huawei Cloud to create a global leading Internet of Things big data platform for structural health monitoring and protect project safety with artificial intelligence and other new technologies. Deep learning image recognition module based on ModelArts realizes automatic monitoring of construction standardization.
Technology benefits life, ModelArts makes AI development more efficient. Facing the future, Huawei Cloud will continue to optimize ModelArts, create a better enterprise AI development platform, and promote the development of artificial intelligence industry. At the same time, we will continue to expand the developer ecological community. We hope to join hands with more developers and corporate partners to let AI be deployed in more scenes and benefit all industries.
Click to follow, the first time to learn about Huawei cloud fresh technology ~