annyyhyang
/
sam_amp
- Public
- 38.1K runs
Run annyyhyang/sam_amp with an API
Use one of our client libraries to get started quickly. Clicking on a library will take you to the Playground tab where you can tweak different inputs, see the results, and copy the corresponding code to use in your own project.
Input schema
The fields you can use to run this model with an API. If you don't give a value for a field its default value will be used.
| Field | Type | Default value | Description |
|---|---|---|---|
| image |
string
|
Input image
|
|
| mask_limit |
integer
|
maximum number of masks to return. If -1 or None, all masks will be returned. NOTE: The masks are sorted by predicted_iou.
|
|
| points_per_side |
integer
|
32
|
The number of points to be sampled along one side of the image. The total number of points is points_per_side**2. If None, point_grids must provide explicit point sampling.
|
| pred_iou_thresh |
number
|
0.88
|
A filtering threshold in [0,1], using the model's predicted mask quality.
|
| stability_score_thresh |
number
|
0.95
|
A filtering threshold in [0,1], using the stability of the mask under changes to the cutoff used to binarize the model's mask predictions.
|
| stability_score_offset |
number
|
1
|
The amount to shift the cutoff when calculated the stability score.
|
| box_nms_thresh |
number
|
0.7
|
The box IoU cutoff used by non-maximal suppression to filter duplicate masks.
|
| crop_n_layers |
integer
|
0
|
If >0, mask prediction will be run again on crops of the image. Sets the number of layers to run, where each layer has 2**i_layer number of image crops
|
| crop_nms_thresh |
number
|
0.7
|
The box IoU cutoff used by non-maximal suppression to filter duplicate masks between different crops.
|
| crop_overlap_ratio |
number
|
0.3413333333333333
|
Sets the degree to which crops overlap. In the first crop layer, crops will overlap by this fraction of the image length. Later layers with more crops scale down this overlap.
|
| crop_n_points_downscale_factor |
integer
|
1
|
The number of points-per-side sampled in layer n is scaled down by crop_n_points_downscale_factor**n.
|
| min_mask_region_area |
integer
|
0
|
If >0, postprocessing will be applied to remove disconnected regions and holes in masks with area smaller than min_mask_region_area.
|
{
"type": "object",
"title": "Input",
"required": [
"image"
],
"properties": {
"image": {
"type": "string",
"title": "Image",
"format": "uri",
"x-order": 0,
"description": "Input image"
},
"mask_limit": {
"type": "integer",
"title": "Mask Limit",
"x-order": 1,
"description": "maximum number of masks to return. If -1 or None, all masks will be returned. NOTE: The masks are sorted by predicted_iou."
},
"crop_n_layers": {
"type": "integer",
"title": "Crop N Layers",
"default": 0,
"x-order": 7,
"description": "If >0, mask prediction will be run again on crops of the image. Sets the number of layers to run, where each layer has 2**i_layer number of image crops"
},
"box_nms_thresh": {
"type": "number",
"title": "Box Nms Thresh",
"default": 0.7,
"x-order": 6,
"description": "The box IoU cutoff used by non-maximal suppression to filter duplicate masks."
},
"crop_nms_thresh": {
"type": "number",
"title": "Crop Nms Thresh",
"default": 0.7,
"x-order": 8,
"description": "The box IoU cutoff used by non-maximal suppression to filter duplicate masks between different crops."
},
"points_per_side": {
"type": "integer",
"title": "Points Per Side",
"default": 32,
"x-order": 2,
"description": "The number of points to be sampled along one side of the image. The total number of points is points_per_side**2. If None, point_grids must provide explicit point sampling."
},
"pred_iou_thresh": {
"type": "number",
"title": "Pred Iou Thresh",
"default": 0.88,
"x-order": 3,
"description": "A filtering threshold in [0,1], using the model's predicted mask quality."
},
"crop_overlap_ratio": {
"type": "number",
"title": "Crop Overlap Ratio",
"default": 0.3413333333333333,
"x-order": 9,
"description": "Sets the degree to which crops overlap. In the first crop layer, crops will overlap by this fraction of the image length. Later layers with more crops scale down this overlap."
},
"min_mask_region_area": {
"type": "integer",
"title": "Min Mask Region Area",
"default": 0,
"x-order": 11,
"description": "If >0, postprocessing will be applied to remove disconnected regions and holes in masks with area smaller than min_mask_region_area."
},
"stability_score_offset": {
"type": "number",
"title": "Stability Score Offset",
"default": 1,
"x-order": 5,
"description": "The amount to shift the cutoff when calculated the stability score."
},
"stability_score_thresh": {
"type": "number",
"title": "Stability Score Thresh",
"default": 0.95,
"x-order": 4,
"description": "A filtering threshold in [0,1], using the stability of the mask under changes to the cutoff used to binarize the model's mask predictions."
},
"crop_n_points_downscale_factor": {
"type": "integer",
"title": "Crop N Points Downscale Factor",
"default": 1,
"x-order": 10,
"description": "The number of points-per-side sampled in layer n is scaled down by crop_n_points_downscale_factor**n."
}
}
}
Output schema
The shape of the response you’ll get when you run this model with an API.
Schema
{
"type": "array",
"items": {
"type": "string",
"format": "uri"
},
"title": "Output"
}