kornia/mobile_sam

Pretrained weights for MobileSAM, used by kornia.models.SegmentAnything.

MobileSAM replaces the heavy ViT-H image encoder of SAM with a distilled TinyViT-5M encoder, reducing model size by 60× while retaining most of SAM's segmentation quality. arXiv 2023.

Original repo: ChaoningZhang/MobileSAM

Weights

File Description
mobile_sam.pt TinyViT-5M encoder + SAM decoder

Citation

@article{zhang2023mobilesam,
    title   = {Faster Segment Anything: Towards Lightweight SAM
               for Mobile Applications},
    author  = {Zhang, Chaoning and Han, Dongshen and Qiao, Yu
               and Kim, Jung Uk and Bae, Sung-Ho and Lee, Seungkyu
               and Hong, Choong Seon},
    journal = {arXiv preprint arXiv:2306.14289},
    year    = {2023}
}
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Paper for kornia/mobile_sam