tfm.vision.layers.TemporalSoftmaxPool

Creates a network layer corresponding to temporal softmax pooling.

tfm.vision.layers.TemporalSoftmaxPool(
    trainable=True, name=None, dtype=None, dynamic=False, **kwargs
)

This is useful for multi-class logits (used in e.g., Charades). Modified from AssembleNet Charades evaluation from:

Michael S. Ryoo, AJ Piergiovanni, Mingxing Tan, Anelia Angelova. AssembleNet: Searching for Multi-Stream Neural Connectivity in Video Architectures. (https://arxiv.org/pdf/1905.13209.pdf).

Methods

call

View source

call(
    inputs
)

Calls the layer with the given inputs.