add resnest and xception65

README.md

@@ -1,3 +1,10 @@
+# new features:
+## you can use Resnest and xception as Backbone!(xception65 without pretrained)
+- resnest50
+- resnest101
+- resnest200
+- resnest269
+- xception65
 # FastFCN: Rethinking Dilated Convolution in the Backbone for Semantic Segmentation
 [[Project]](http://wuhuikai.me/FastFCNProject/)    [[Paper]](http://wuhuikai.me/FastFCNProject/fast_fcn.pdf)    [[arXiv]](https://arxiv.org/abs/1903.11816)    [[Home]](http://wuhuikai.me) 
 

encoding/dilated/__init__.py

@@ -1,2 +1,4 @@
 """Dilated ResNet and DenseNet"""
 from .resnet import *
+from .resnest import *
+from .xception import *

encoding/dilated/resnest.py

@@ -0,0 +1,77 @@
+##+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+## Created by: Hang Zhang
+## Email: [email protected]
+## Copyright (c) 2020
+##
+## LICENSE file in the root directory of this source tree 
+##+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+"""ResNeSt models"""
+
+import torch
+from .resnet import ResNet, Bottleneck
+from ..models.model_store import get_model_file
+
+__all__ = ['resnest50', 'resnest101', 'resnest200', 'resnest269']
+
+_url_format = 'https://hangzh.s3.amazonaws.com/encoding/models/{}-{}.pth'
+
+
+def resnest50(pretrained=False, root='~/.encoding/models', **kwargs):
+    model = ResNet(Bottleneck, [3, 4, 6, 3],
+                   radix=2, groups=1, bottleneck_width=64,
+                   deep_stem=True, stem_width=32, avg_down=True,
+                   avd=True, avd_first=False, **kwargs)
+    if pretrained:
+        model.load_state_dict(torch.load(
+            get_model_file('resnest50', root=root)), strict=True)
+    return model
+
+def resnest101(pretrained=False, root='~/.encoding/models', **kwargs):
+    model = ResNet(Bottleneck, [3, 4, 23, 3],
+                   radix=2, groups=1, bottleneck_width=64,
+                   deep_stem=True, stem_width=64, avg_down=True,
+                   avd=True, avd_first=False, **kwargs)
+    if pretrained:
+        model.load_state_dict(torch.load(
+            get_model_file('resnest101', root=root)), strict=True)
+    return model
+
+def resnest200(pretrained=False, root='~/.encoding/models', **kwargs):
+    model = ResNet(Bottleneck, [3, 24, 36, 3],
+                   radix=2, groups=1, bottleneck_width=64,
+                   deep_stem=True, stem_width=64, avg_down=True,
+                   avd=True, avd_first=False, **kwargs)
+    if pretrained:
+        model.load_state_dict(torch.load(
+            get_model_file('resnest200', root=root)), strict=False)
+    return model
+
+def resnest269(pretrained=False, root='~/.encoding/models', **kwargs):
+    model = ResNet(Bottleneck, [3, 30, 48, 8],
+                   radix=2, groups=1, bottleneck_width=64,
+                   deep_stem=True, stem_width=64, avg_down=True,
+                   avd=True, avd_first=False, **kwargs)
+    if pretrained:
+        model.load_state_dict(torch.load(
+            get_model_file('resnest269', root=root)), strict=True)
+    return model
+
+def resnest50_fast(pretrained=False, root='~/.encoding/models', **kwargs):
+    model = ResNet(Bottleneck, [3, 4, 6, 3],
+                   radix=2, groups=1, bottleneck_width=64,
+                   deep_stem=True, stem_width=32, avg_down=True,
+                   avd=True, avd_first=True, **kwargs)
+    if pretrained:
+        model.load_state_dict(torch.load(
+            get_model_file('resnest50fast', root=root)), strict=True)
+    return model
+
+def resnest101_fast(pretrained=False, root='~/.encoding/models', **kwargs):
+    model = ResNet(Bottleneck, [3, 4, 23, 3],
+                   radix=2, groups=1, bottleneck_width=64,
+                   deep_stem=True, stem_width=64, avg_down=True,
+                   avd=True, avd_first=True, **kwargs)
+    if pretrained:
+        model.load_state_dict(torch.load(
+            get_model_file('resnest101fast', root=root)), strict=True)
+    return model