sk_resnet_18

→SKNet

sk_resnet_18(pretrained: bool = False, overrides: object = {})

source edit

SK-ResNet-18 feature-extracting backbone (no classification head).

Builds an SKNet with ResNet-18 topology (_SelectiveKernelBasic blocks stacked [2, 2, 2, 2]). Both $3 \times 3$ convolutions inside every block are replaced by Selective Kernel units, giving full SK treatment of the basic-block design. Approximately 24.7M parameters.

Model Size

Parameters

pretrainedbool= False

Reserved for future pretrained-weight loading. Currently ignored — the returned model is randomly initialised.

**overridesobject= {}

Keyword overrides forwarded into SKNetConfig.

Returns

SKNet

Backbone with the SK-ResNet-18 configuration applied (or with overrides merged on top of it).

Notes

See Li et al., "Selective Kernel Networks", CVPR 2019 (arXiv:1903.06586). Uses rd_ratio = 0.6 for the attention bottleneck and split_input = False so each branch receives the full input — matching the SK-ResNet-18 budget reference.

Examples

>>> import lucid
>>> from lucid.models.vision.sknet import sk_resnet_18
>>> model = sk_resnet_18()
>>> x = lucid.randn(1, 3, 224, 224)
>>> out = model(x)
>>> out.last_hidden_state.shape
(1, 512, 7, 7)

Used by 1

lucid.models.vision.sknet

>>> import lucid >>> from lucid.models.vision.sknet import sk_resnet_18 >>> model = sk_resnet_18() >>> x = lucid.randn(1, 3, 224, 224) >>> out = model(x) >>> out.last_hidden_state.shape (1, 512, 7, 7)