#199 · Primary category： Computer Vision
efficientsam3
EfficientSAM3 compresses SAM3 into lightweight, edge-friendly models via progressive knowledge distillation for fast promptable concept segmentation and tracking.
Project last updated：08/11/26
GitHub Stars
658
Forks
53
Contributors
6
License
Apache-2.0
Why we included this project
EfficientSAM3 takes the full SAM3 model and shrinks it into a family of smaller student models through progressive knowledge distillation, so text-prompted segmentation and video tracking can run on phones, embedded boards, and modest GPUs. The released checkpoints pair several backbone choices, including RepViT, TinyViT, and EfficientViT, with lightweight MobileCLIP text encoders, letting you trade speed against accuracy for your target device. Training and evaluation scripts come with dataset preparation for COCO, DAVIS, LVOS, and YouTube-VOS, which is useful if you want to fine-tune the weights rather than just load pretrained ones. Teams building edge vision products around segmentation and tracking get a documented, reproducible route from the research model to deployable lightweight weights.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
opencv
Open Source Computer Vision Library
RuView
π RuView turns commodity WiFi signals into real-time spatial intelligence, vital sign monitoring, and presence detection — all without a single pixel of video.
PaddleOCR
Turn any PDF or image document into structured data for your AI. A powerful, lightweight OCR toolkit that bridges the gap between images/PDFs and LLMs. Supports 100+ languages.
MinerU
Transforms complex documents like PDFs and Office docs into LLM-ready markdown/JSON for your Agentic workflows.
tesseract
Tesseract Open Source OCR Engine (main repository)