Hugging Face's logo

xiao-1011
/
dmc-brainmap-registration-predictor

Keypoint Detection
PyTorch
computer-vision
image-registration
homography-estimation
paired-image
neuroscience
histology
brain-atlas
napari
Model card Files
xet
 Community
dmc-brainmap-registration-predictor
File size: 3,027 Bytes
c2021af
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
{
  "model_version": "1.0.0",
  "data": {
    "species": "Mus musculus",
    "strain": "C57BL/6J",
    "sexes": ["female", "male"],
    "age_group": "adult",
    "imaging": "epifluorescence microscopy",
    "sample_channel": "green channel, EGFP spectral range",
    "atlas_id": "allen_mouse_10um",
    "atlas_version": "1.2",
    "atlas_resolution_um": 10.0,
    "image_width_px": 1140,
    "image_height_px": 800,
    "total_animals": 87,
    "total_sections": 5613,
    "split_unit": "animal",
    "split_seed": 42,
    "splits": {
      "train": {"animals": 61, "sections": 4015},
      "validation": {"animals": 13, "sections": 745},
      "test": {"animals": 13, "sections": 853}
    },
    "target_registration_software": "SHARPy-track within DMC-BrainMap",
    "target_reviewers": ["Felix Jung", "Xiao Cao", "Loran Heymans"],
    "acceptable_mean_tre_px": 4.0,
    "acceptable_mean_tre_um": 40.0,
    "maximum_landmark_homography_fit_rmse_px": 60.0,
    "excluded_above_fit_rmse": 296,
    "training_images_public": false
  },
  "model": {
    "model_name": "spatial",
    "base_channels": 24,
    "spatial_pool_size": [5, 8],
    "spatial_hidden_channels": 256,
    "max_corner_offset_px": 512.0,
    "parameter_count": 2336200
  },
  "optimization": {
    "maximum_epochs": 200,
    "stopped_after_epoch": 111,
    "selected_epoch": 91,
    "selection_metric": "validation_loss",
    "selected_validation_loss": 16.35543269727054,
    "batch_size": 4,
    "optimizer": "AdamW",
    "learning_rate": 0.0001,
    "weight_decay": 0.0001,
    "scheduler": "ReduceLROnPlateau",
    "scheduler_factor": 0.5,
    "scheduler_patience": 6,
    "minimum_learning_rate": 5e-7,
    "early_stopping_patience": 20,
    "early_stopping_min_delta": 0.03,
    "seed": 42
  },
  "loss": {
    "primary": "Smooth L1 reprojection loss",
    "points_per_axis": 8,
    "sample_grid_points": 64,
    "sample_bbox_inset_fraction": 0.02,
    "corner_loss": "Smooth L1",
    "corner_loss_weight": 0.25
  },
  "augmentation": {
    "applied_splits": ["train"],
    "preset": "moderate",
    "seed": 42,
    "changes_geometry": false,
    "missing_chunk_probability": 0.45,
    "missing_chunk_count": [1, 3],
    "missing_chunk_area_fraction": [0.02, 0.10],
    "tear_probability": 0.35,
    "crack_probability": 0.30,
    "fold_probability": 0.30,
    "exposure_probability": 0.65,
    "brightness_delta": [-22.0, 22.0],
    "contrast_factor": [0.75, 1.35],
    "gamma": [0.70, 1.45],
    "noise_sigma": [0.0, 8.0]
  },
  "environment": {
    "python": "3.14",
    "pytorch": "2.12.0+cu130",
    "torchvision": "0.27.0+cu130",
    "numpy": "2.4.4",
    "opencv_python": "4.13.0.92",
    "gpu": "NVIDIA GeForce RTX 3070 Ti"
  },
  "provenance": {
    "source_repository": "https://github.com/xiao-1011/sharpy-model",
    "source_commit": "0a80f4f831f6450a015a2f32433f22975ac97dc0",
    "run_name": "baseline-005-spatial-augmented-200ep",
    "original_checkpoint_sha256": "2f6c84b12a8449a63ff45973cb91a81326973915d4801659c99b973087c1dac8"
  }
}