def r2plus1d_34(num_classes, pretrained=False, progress=False, arch=None):
model = VideoResNet(block=BasicBlock,
conv_makers=[Conv2Plus1D] * 4,
layers=[3, 4, 6, 3],
stem=R2Plus1dStem)
model.fc = nn.Linear(model.fc.in_features, out_features=num_classes)
# Fix difference in PyTorch vs Caffe2 architecture
# https://github.com/facebookresearch/VMZ/issues/89
# https://github.com/pytorch/vision/issues/1265
model.layer2[0].conv2[0] = Conv2Plus1D(128, 128, 288)
model.layer3[0].conv2[0] = Conv2Plus1D(256, 256, 576)
model.layer4[0].conv2[0] = Conv2Plus1D(512, 512, 1152)
# We need exact Caffe2 momentum for BatchNorm scaling
for m in model.modules():
if isinstance(m, nn.BatchNorm3d):
m.eps = 1e-3
m.momentum = 0.9
if pretrained:
state_dict = torch.hub.load_state_dict_from_url(model_urls[arch],
progress=progress)
model.load_state_dict(state_dict)
return model
def r3d_34(cfg):
model = VideoResNet(block=BasicBlock,
conv_makers=[Conv3DSimple] * 4,
layers=[3, 4, 6, 3],
stem=BasicStem)
model.fc = nn.Linear(model.fc.in_features, cfg.CONFIG.DATA.NUM_CLASSES)
return model
def r3d_18(cfg):
model = VideoResNet(block=BasicBlock,
conv_makers=[Conv3DSimple] * 4,
layers=[2, 2, 2, 2],
stem=BasicStem)
state_dict = load_state_dict_from_url(model_urls['r3d_18'], progress=True)
model.load_state_dict(state_dict)
if cfg.CONFIG.DATA.DATASET != 'kinetics':
model.fc = nn.Linear(model.fc.in_features, cfg.CONFIG.DATA.NUM_CLASSES)
print('Pretrained Model Weight Loaded')
return model
def _generic_resnet(arch, pretrained=False, progress=False, **kwargs):
model = VideoResNet(**kwargs)
# We need exact Caffe2 momentum for BatchNorm scaling
for m in model.modules():
if isinstance(m, nn.BatchNorm3d):
m.eps = 1e-3
m.momentum = 0.9
if pretrained:
state_dict = torch.hub.load_state_dict_from_url(model_urls[arch],
progress=progress)
model.load_state_dict(state_dict)
return model
def r2plus1d_34(cfg):
model = VideoResNet(block=BasicBlock,
conv_makers=[Conv2Plus1D] * 4,
layers=[3, 4, 6, 3],
stem=R2Plus1dStem)
model.layer2[0].conv2[0] = Conv2Plus1D(128, 128, 288)
model.layer3[0].conv2[0] = Conv2Plus1D(256, 256, 576)
model.layer4[0].conv2[0] = Conv2Plus1D(512, 512, 1152)
model.fc = nn.Linear(model.fc.in_features, 359)
state_dict = torch.load('./logs/r2plus1d/pretrained.pth')
model.load_state_dict(state_dict)
model.fc = nn.Linear(model.fc.in_features, cfg.CONFIG.DATA.NUM_CLASSES)
print('Pretrained Model Weight Loaded')
return model
def r2plus1d_34(pretrained=True, progress=False, **kwargs):
model = VideoResNet(
block=BasicBlock,
conv_makers=[Conv2Plus1D] * 4,
layers=[3, 4, 6, 3],
stem=R2Plus1dStem,
**kwargs,
)
# We need exact Caffe2 momentum for BatchNorm scaling
for m in model.modules():
if isinstance(m, nn.BatchNorm3d):
m.eps = 1e-3
m.momentum = 0.9
if pretrained:
state_dict = torch.hub.load_state_dict_from_url(R2PLUS1D_34_MODEL_URL,
progress=progress)
model.load_state_dict(state_dict)
return model
def r2plus1d_34(num_classes):
model = VideoResNet(block=BasicBlock,
conv_makers=[Conv2Plus1D] * 4,
layers=[3, 4, 6, 3],
stem=R2Plus1dStem)
model.fc = nn.Linear(model.fc.in_features, out_features=num_classes)
# Fix difference in PyTorch vs Caffe2 architecture
# https://github.com/facebookresearch/VMZ/issues/89
model.layer2[0].conv2[0] = Conv2Plus1D(128, 128, 288)
model.layer3[0].conv2[0] = Conv2Plus1D(256, 256, 576)
model.layer4[0].conv2[0] = Conv2Plus1D(512, 512, 1152)
# We need exact Caffe2 momentum for BatchNorm scaling
for m in model.modules():
if isinstance(m, nn.BatchNorm3d):
m.eps = 1e-3
m.momentum = 0.9
return model
def r2plus1d_34(cfg):
model = VideoResNet(block=BasicBlock,
conv_makers=[Conv2Plus1D] * 4,
layers=[3, 4, 6, 3],
stem=R2Plus1dStem)
model.layer2[0].conv2[0] = Conv2Plus1D(128, 128, 288)
model.layer3[0].conv2[0] = Conv2Plus1D(256, 256, 576)
model.layer4[0].conv2[0] = Conv2Plus1D(512, 512, 1152)
model.fc = nn.Linear(model.fc.in_features, 359)
state_dict = torch.load('./logs/r2plus1d/pretrained.pth')
model.load_state_dict(state_dict)
model.fc = nn.Linear(model.fc.in_features, cfg.CONFIG.DATA.NUM_CLASSES)
print('Pretrained Model Weight Loaded')
return model
if __name__ == '__main__':
model = VideoResNet(block=BasicBlock,
conv_makers=[Conv3DSimple] * 4,
layers=[2, 2, 2, 2],
stem=BasicStem)
state_dict = load_state_dict_from_url(model_urls['r3d_18'], progress=True)
model.load_state_dict(state_dict)
model.fc = nn.Linear(model.fc.in_features, 101)
print('Pretrained Model Weight Loaded')
input = torch.randn((2, 3, 4, 224, 224))
output = model(input)
print(output.shape)