August 2020 Summaries

Clarifai Release 6.7 introduces several innovations to streamline AI activities into a cohesive lifecycle, enhancing the scalability and maintainability of AI solutions. It offers dynamic model types that simplify accessing, sorting, and filtering models based on various criteria such as model type, inputs, and outputs, thus facilitating the integration of different model layers and types for desired outcomes. The AWS Lambda Model allows seamless integration of Clarifai functionalities with AWS Lambda, enabling the development and management of application functionalities without the need for extensive infrastructure, and simplifying access to images stored in S3 buckets. Additionally, the release emphasizes security with the implementation of two-factor authentication (2FA), which provides significant protection against cyber threats, and is particularly crucial for compliance in enterprise environments, although beneficial for all organizations concerned with account and data safety.

Clarifai has launched version 6.6 of its platform, featuring the new Labeler tool designed to facilitate efficient and precise data labeling for AI-driven projects. Labeler offers enhanced features such as bounding box and polygon labeling for improved object detection and semantic segmentation, as well as video interpolation for tracking movement across frames. Image editing tools and task management features are included to support professional data annotators in scaling labeling operations and managing team performance. Additionally, the platform introduces Collectors, which streamline data ingestion and management by automatically integrating input data from production models, thus enabling active learning and other advanced training capabilities. Collectors and their functionalities are available in both Essential and Enterprise plans, allowing users to monitor specific models and customize data sampling rules.