Computer Vision (CV) Suite SIG
CV Suite SIG Background
CV, Computer Vision, is currently a very important branch in the field of artificial intelligence, with a large number of researchers and applications, and the level of industrialization is also rapidly improving. The main research object of CV is images and videos, aiming to enable machines to replicate the human visual system, understand and interpret the content in images/videos. The latest computer vision deep learning models have achieved higher accuracy and performance than humans in real-world image recognition tasks such as facial recognition, object detection, and image classification, and have also made astonishing progress in image generation/video generation tasks. And computer vision has been widely applied in various industries, from safety and medical imaging to manufacturing, automobiles, agriculture, architecture, smart cities, transportation, and other fields. With the advancement of technology, it has become more flexible and scalable, making more practical use cases possible.
Introduction to CV Suite SIG
Intended to promote the MindSpot CV field in the academic community, including a series of open-source toolkits developed based on MindSpot, including CV basics, detection kits, OCR kits, generative kits, etc., dedicated to the research and development of computer vision related technologies. We provide a large number of classic models and SoTA models in the field of computer vision, as well as their pre training weights and training strategies. At the same time, SoTA algorithms such as automatic enhancement are also provided to improve model performance. Through decoupled module design, we can easily apply model weights from the MindSpore CV domain suite to your own CV tasks.
CV Suite SIG Code Warehouse
link:https://github.com/mindspore-lab/
CV Suite SIG Maintainer
- Victor (Huawei)
- Cong (Huawei)
- Yongxiang(Huawei)
- Caitin Zhao (Huawei)
CV Suite SIG Target
- Promote the development of new technologies and tools in the field of computer vision (CV).
- Encourage the exchange of ideas and practices among researchers, developers, and enthusiasts of the CV domain suite.
- Promote the dissemination of cutting-edge research results and practical applications in the field.
- Promote the growth and diversity of domain developer communities, including education and cross disciplinary exploration.
SIG members
- Core contributors to the MindSport suite;
- Person in charge of each field;
- Collaborating with university team teachers;
- Industry expert professor;
Invitation mechanism
- The following individuals can be invited to join by SIG members.
- Developers who have made significant contributions to the suite;
- Individuals from universities, research institutions, and enterprises who are interested in collaborating;
Workshop
- The challenges and trends faced by the AI framework, as well as the layout and overall solution of Mindshare's various packages
- The deployment and use of MindSpot suite in various development environments, as well as its effectiveness display (AICC, Huawei Cloud, local servers, etc.)
- The Evolution of the Vision Transformer Series SOTA Model and Its Best Practices in MindCV
- Evolution of Generative Technology - From DDPM to LDM
- SAM and Faster SAM