SimpleDet Docs

Examples

Three practical detector patterns that match the current suite plus native runtime API.

Overview Quickstart Core Concepts API Reference

Use this page when you want a concrete starting point for a detector family instead of a generic tutorial.

Satellite imagery vessel detection

from simpledet.suite import build_detector, build_neck

detector_spec = build_detector(
    "retinanet",
    encoder="convnext_tiny.in12k_ft_in1k",
    neck=build_neck(name="FPN", out_channels=192, num_outs=5),
    num_classes=1,
    in_channels=4,
)

Notebook starting point: notebooks/Tools/suite_01_dense_detector.ipynb.

Building detection

from simpledet.suite import build_detector, build_neck

detector_spec = build_detector(
    "faster_rcnn",
    encoder="convnext_tiny.in12k_ft_in1k",
    neck=build_neck(name="FPN", out_channels=192),
    num_classes=1,
    in_channels=3,
)

Notebook starting point: notebooks/Tools/suite_02_roi_detector.ipynb.

Ship detection with FCOS

from simpledet.suite import build_detector, build_neck

detector_spec = build_detector(
    "fcos",
    encoder="convnext_tiny.in12k_ft_in1k",
    neck=build_neck(name="FPN", out_channels=192, num_outs=4),
    num_classes=1,
    in_channels=3,
)

Notebook starting point: notebooks/Tools/suite_01_dense_detector.ipynb.

API-first training flow

from simpledet import run_training
from simpledet.suite import build_detector

result = run_training(
    dataset_root='/path/to/data',
    detector_spec=build_detector("retinanet", ...),
    categories=('wake',),
    in_channels=3,
    tif_channels_to_load=[1, 2, 3],
)

Notebook starting point: notebooks/Tools/api_pipeline_training_example.ipynb.

API-first inference flow

from simpledet import run_inference
from simpledet.suite import build_detector

spec = build_detector("fcos", ...)
result = run_inference(
    dataset_root='/path/to/data',
    categories=('wake',),
    in_channels=3,
    detector_spec=spec,
    tif_channels_to_load=[1, 2, 3],
)

Notebook starting point: notebooks/Tools/api_inference_example.ipynb.