Ianvs is a distributed synergy AI benchmarking project incubated in KubeEdge SIG AI. Ianvs aims to test the performance of distributed synergy AI solutions following recognized standards, in order to facilitate more efficient and effective development. More detailedly, Ianvs prepares not only test cases with datasets and corresponding algorithms, but also benchmarking tools including simulation and hyper-parameter searching. Ianvs also reveals best practices for developers and end users with presentation tools including leaderboards and test reports.
The distributed synergy AI benchmarking Ianvs aims to test the performance of distributed synergy AI solutions following recognized standards, in order to facilitate more efficient and effective development.
The scope of Ianvs includes
The architectures and related concepts are shown in the below figure. The ianvs is designed to run within a single node. Critical components include
More details on Ianvs components:
Documentation is located on readthedoc.io. The documents include the quick start, guides, dataset descriptions, algorithms, user interfaces, stories, and roadmap.
Follow the Ianvs installation document to install Ianvs.
Scenario PCB-AoI:Industrial Defect Detection on the PCB-AoI Dataset.
Example PCB-AoI-1:Testing single task learning in industrial defect detection.
Example PCB-AoI-2:Testing incremental learning in industrial defect detection.
Scenario Cityscapes-Synthia: Curb Detetion on Cityscapes-Synthia Dataset
Example Cityscapes-Synthia-1: Lifelong learning in semantic segmentation
Example Cityscapes-Synthia-2: Lifelong learning in curb detetion
Example Cityscapes-Synthia-3: Scene based unknown task recognition in curb detetion
Example Cityscapes-Synthia-4: Integrating GAN and Self-taught Learning into Ianvs Lifelong Learning
Scenario Cloud-Robotics: Semantic Segmentation on Cloud-Robotics Dataset
Example Cloud-Robotics-1: Lifelong learning in semantic segmentation
Example Cloud-Robotics-2: Class increment learning in semantic segmentation
Example Cloud-Robotics-3: Lifelong learning in sam annotation
Routine Community Meeting for KubeEdge SIG AI runs weekly:
Resources:
If you have questions, feel free to reach out to us in the following ways:
If you're interested in being a contributor and want to get involved in developing the Ianvs code, please see CONTRIBUTING for details on submitting patches and the contribution workflow.
Ianvs is under the Apache 2.0 license. See the LICENSE file for details.