facebook/sam2.1-hiera-tiny is a compact open-weight model from Meta's Segment Anything Model 2 family. It supports promptable mask generation for both images and videos, including point clicks, bounding boxes, and multi-object segmentation. The model is designed for easy integration via the Hugging Face Transformers library and runs locally for computer vision research and applications.
Sam2.1 Hiera Tiny is an Other AI project. It focuses on performing accurate object segmentation in images and videos using prompt-based inputs. Sam2.1 Hiera Tiny is an open-source project aimed at developers. Sam2.1 Hiera Tiny is open source under the Apache-2.0 license. It runs on the web and API.
It is developed by Meta AI (United States), and it first shipped in 2024. The project is developed in the open on GitHub with 19.6k stars. Key capabilities include mask generation, video segmentation, and point prompts.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do