Skip to content

Use FX feature extractor for segm model #4563

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Oct 7, 2021
Merged

Use FX feature extractor for segm model #4563

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
9258054
Use FX feature extractor for segm model
vfdev-5 Oct 7, 2021
21b0ff8
Removed use_fe option
vfdev-5 Oct 7, 2021
ad07e8c
Merge branch 'main' of github.com:pytorch/vision into segm-model-with-fe
vfdev-5 Oct 7, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions torchvision/models/segmentation/segmentation.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,7 @@
from ..._internally_replaced_utils import load_state_dict_from_url
from .. import mobilenetv3
from .. import resnet
from .._utils import IntermediateLayerGetter
from ..feature_extraction import create_feature_extractor
from .deeplabv3 import DeepLabHead, DeepLabV3
from .fcn import FCN, FCNHead
from .lraspp import LRASPP
Expand Down Expand Up @@ -60,7 +60,7 @@ def _segm_model(
return_layers = {out_layer: "out"}
if aux:
return_layers[aux_layer] = "aux"
backbone = IntermediateLayerGetter(backbone, return_layers=return_layers)
backbone = create_feature_extractor(backbone, return_layers)

aux_classifier = None
if aux:
Expand Down Expand Up @@ -116,7 +116,7 @@ def _segm_lraspp_mobilenetv3(backbone_name: str, num_classes: int, pretrained_ba
low_channels = backbone[low_pos].out_channels
high_channels = backbone[high_pos].out_channels

backbone = IntermediateLayerGetter(backbone, return_layers={str(low_pos): "low", str(high_pos): "high"})
backbone = create_feature_extractor(backbone, {str(low_pos): "low", str(high_pos): "high"})

model = LRASPP(backbone, low_channels, high_channels, num_classes)
return model
Expand Down