models.json

{
  "US/USF-MAE.pt": {
    "fov": {
      "formula": "Global(ViT-B/16, stride 16)"
    },
    "dimension": "2",
    "numberofchannels": "3",
    "layers": 4,
    "license": "MIT",
    "patch_size": null,
    "size_multiple": 16,
    "source": {
      "name": "USF-MAE",
      "code": "https://github.com/Yusufii9/USF-MAE",
      "weights": "https://drive.google.com/file/d/1ZPu_7KhMEuaq-XdLhVp2EEgMgLJ4dKhr/view",
      "paper": "https://doi.org/10.1016/j.bspc.2026.110313",
      "architecture": "MAE ViT-B/16 encoder, blocks 2/5/8/11 tapped as stride-16 feature maps",
      "checkpoint_sha256": "f815c629878c17136985af9f4fdc81c2cfa02a94e4d992c026699957f75ccb66"
    }
  },
  "US/SAMUS.pt": {
    "fov": {
      "formula": "Global(256 px tile, stride 8)"
    },
    "dimension": "2",
    "numberofchannels": "1",
    "layers": 5,
    "license": "MIT",
    "patch_size": 256,
    "size_multiple": null,
    "source": {
      "name": "SAMUS",
      "code": "https://github.com/xianlin7/SAMUS",
      "weights": "https://drive.google.com/file/d/1nQjMAvbPeolNpCxQyU_HTiOiB5704pkH/view",
      "paper": "https://doi.org/10.48550/arXiv.2309.06824",
      "architecture": "SAM ViT-B image encoder at 256 px, patch size 8, global-attention blocks 2/5/8/11 plus the 256-channel neck",
      "checkpoint_sha256": "a95de25d7b9065a92e268d73b1e76b871d7b0e4e66ce872b1e298f5b6c446fc4"
    }
  }
}