Ready to use out of the box . Open source datasets and 10+ supporting algorithms cover the entire process of preprocessing, pretraining, training, reasoning, and post-processing, ready to use out of the box without any modification.
Flexible switching across all scenarios : Test case management unifies different scenarios and their AI algorithm architecture and interfaces, and can use a set of tools to be compatible with multiple AI paradigms such as incremental, lifelong, and federated.
Extensible and open tool chain : Test environment management enables customized dynamic configuration of test data sets and test indicators, saying goodbye to closed and conservative test environments.
Generate test cases with low code : Test case management generates kuwait mobile phone number list test cases based on grid search and other auxiliary methods. For example, one configuration file can implement multiple hyperparameter tests, reducing the tedious repetitive programming during hyperparameter search. Ianvs won the honor of Outstanding Cloud-Edge X Project at the Cloud-Edge Collaboration Conference 2022.
Figure 3 KubeEdge-Ianvs overall architecture
2. Cloud-native edge intelligence application cases
This article starts from 20+ project cases in 7 scenarios, and shares the experience of industrial application of cloud-native edge intelligence based on the standardized unified framework KubeEdge-Sedna and the benchmark KubeEdge-Ianvs of the KubeEdge community SIG AI ecosystem.
Pedestrian trajectory tracking series of cases . For example, in security scenarios, managers need to know the specific location of specific park personnel at a specific time. Traditional security systems require security personnel to search image by image based on personal information under real-time monitoring screens or hundreds of hours of surveillance videos, which requires a lot of manpower and is inefficient.