lvis-api
The LVIS API facilitates effective interaction with a large-scale dataset containing over 2 million instance segmentation masks within more than 1200 object categories. It offers tools for annotation manipulation, visualization, and result evaluation. This API, supporting Large Vocabulary Instance Segmentation, is essential for advancing computer vision research. The publicly available v1.0 release is integral to the LVIS Challenge at ECCV 2020. Install the API effortlessly in a virtual environment to seamlessly integrate with the COCO toolkit and boost research effectiveness.
