neodroidvision.mixed.architectures.self_attention_network.self_attention_modules.functional.subtraction

neodroidvision.mixed.architectures.self_attention_network.self_attention_modules.functional.subtraction(input: Tensor, kernel_size: int = 3, stride: int = 1, padding: int = 0, dilation: int = 1, pad_mode: PadModeEnum = PadModeEnum.ref_pad) Tensor[source]
Parameters
  • input

  • kernel_size

  • stride

  • padding

  • dilation

  • pad_mode

Returns

Return type