ConvNet

Our implementation of ConvNet is based on DC.

By default, we use width 128, average pooling, and ReLU activation. We provide the following interface to initialize a ConvNet model:

dd_ranking.utils.get_convnet(model_name: str, im_size: tuple, channel: int, num_classes: int, net_depth: int, net_norm: str, pretrained: bool, model_path: str) [SOURCE]

Parameters

  • model_name(str): Name of the model. Please navigate to models for the model naming convention in DD-Ranking.
  • im_size(tuple): Image size.
  • channel(int): Number of channels of the input image.
  • num_classes(int): Number of classes.
  • net_depth(int): Depth of the network.
  • net_norm(str): Normalization method. In ConvNet, we support instance, batch, and group normalization.
  • pretrained(bool): Whether to load pretrained weights.
  • model_path(str): Path to the pretrained model weights.

To load a ConvNet model with different width or activation function or pooling method, you can use the following interface:

dd_ranking.utils.networks.ConvNet(channel, num_classes, net_width, net_depth, net_act, net_norm, net_pooling, im_size) [SOURCE]

Parameters

We only list the parameters that are not present in get_convnet.

  • net_width(int): Width of the network.
  • net_act(str): Activation function. We support relu, leakyrelu, and sigmoid.
  • net_pooling(str): Pooling method. We support avgpooling, maxpooling, and none.