JDwu1120
commented
5 months ago I3Res50( (conv1): Conv3d(3, 64, kernel_size=(5, 7, 7), stride=(2, 2, 2), padding=(2, 3, 3), bias=False) (bn1): BatchNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) (maxpool1): MaxPool3d(kernel_size=(2, 3, 3), stride=(2, 2, 2), padding=(0, 0, 0), dilation=1, ceil_mode=False) (maxpool2): MaxPool3d(kernel_size=(2, 1, 1), stride=(2, 1, 1), padding=(0, 0, 0), dilation=1, ceil_mode=False) (layer1): Sequential( (0): Bottleneck( (conv1): Conv3d(64, 64, kernel_size=(3, 1, 1), stride=(1, 1, 1), padding=(1, 0, 0), bias=False) (bn1): BatchNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(64, 64, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(64, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) (downsample): Sequential( (0): Conv3d(64, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (1): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) ) ) (1): Bottleneck( (conv1): Conv3d(256, 64, kernel_size=(3, 1, 1), stride=(1, 1, 1), padding=(1, 0, 0), bias=False) (bn1): BatchNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(64, 64, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(64, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (2): Bottleneck( (conv1): Conv3d(256, 64, kernel_size=(3, 1, 1), stride=(1, 1, 1), padding=(1, 0, 0), bias=False) (bn1): BatchNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(64, 64, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(64, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) ) (layer2): Sequential( (0): Bottleneck( (conv1): Conv3d(256, 128, kernel_size=(3, 1, 1), stride=(1, 1, 1), padding=(1, 0, 0), bias=False) (bn1): BatchNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(128, 128, kernel_size=(1, 3, 3), stride=(1, 2, 2), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(128, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) (downsample): Sequential( (0): Conv3d(256, 512, kernel_size=(1, 1, 1), stride=(1, 2, 2), bias=False) (1): BatchNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) ) ) (1): Bottleneck( (conv1): Conv3d(512, 128, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn1): BatchNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(128, 128, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(128, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (2): Bottleneck( (conv1): Conv3d(512, 128, kernel_size=(3, 1, 1), stride=(1, 1, 1), padding=(1, 0, 0), bias=False) (bn1): BatchNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(128, 128, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(128, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (3): Bottleneck( (conv1): Conv3d(512, 128, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn1): BatchNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(128, 128, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(128, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) ) (layer3): Sequential( (0): Bottleneck( (conv1): Conv3d(512, 256, kernel_size=(3, 1, 1), stride=(1, 1, 1), padding=(1, 0, 0), bias=False) (bn1): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(256, 256, kernel_size=(1, 3, 3), stride=(1, 2, 2), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(256, 1024, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) (downsample): Sequential( (0): Conv3d(512, 1024, kernel_size=(1, 1, 1), stride=(1, 2, 2), bias=False) (1): BatchNorm3d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) ) ) (1): Bottleneck( (conv1): Conv3d(1024, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn1): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(256, 256, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(256, 1024, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (2): Bottleneck( (conv1): Conv3d(1024, 256, kernel_size=(3, 1, 1), stride=(1, 1, 1), padding=(1, 0, 0), bias=False) (bn1): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(256, 256, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(256, 1024, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (3): Bottleneck( (conv1): Conv3d(1024, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn1): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(256, 256, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(256, 1024, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (4): Bottleneck( (conv1): Conv3d(1024, 256, kernel_size=(3, 1, 1), stride=(1, 1, 1), padding=(1, 0, 0), bias=False) (bn1): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(256, 256, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(256, 1024, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (5): Bottleneck( (conv1): Conv3d(1024, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn1): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(256, 256, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(256, 1024, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) ) (layer4): Sequential( (0): Bottleneck( (conv1): Conv3d(1024, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn1): BatchNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(512, 512, kernel_size=(1, 3, 3), stride=(1, 2, 2), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(512, 2048, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) (downsample): Sequential( (0): Conv3d(1024, 2048, kernel_size=(1, 1, 1), stride=(1, 2, 2), bias=False) (1): BatchNorm3d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) ) ) (1): Bottleneck( (conv1): Conv3d(2048, 512, kernel_size=(3, 1, 1), stride=(1, 1, 1), padding=(1, 0, 0), bias=False) (bn1): BatchNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(512, 512, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(512, 2048, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) (2): Bottleneck( (conv1): Conv3d(2048, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn1): BatchNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv2): Conv3d(512, 512, kernel_size=(1, 3, 3), stride=(1, 1, 1), padding=(0, 1, 1), bias=False) (bn2): BatchNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (conv3): Conv3d(512, 2048, kernel_size=(1, 1, 1), stride=(1, 1, 1), bias=False) (bn3): BatchNorm3d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (relu): ReLU(inplace=True) ) ) (avgpool): AdaptiveAvgPool3d(output_size=(1, 1, 1)) (fc): Linear(in_features=2048, out_features=400, bias=True) (drop): Dropout(p=0.5, inplace=False) ) This is the network structure in the i3D feature extraction library you provided. May I ask which layer of feature output you extracted
Hello author, I downloaded your code and replicated this process using the i3d dataset you provided. The experimental results are basically consistent with what you mentioned in your article, which is really great. However, I also encountered some problems. I used the i3d feature extraction library mentioned in your readme to extract the original dataset you provided, and the model trained was very poor. Moreover, the i3d feature values I extracted are also different from what you provided. I would like to know which layer is the "mix_5c" layer here? Can you upload your code for extracting i3D features?