action-recognition-0001 (composite)

Use Case and High-Level Description

This is a general-purpose action recognition composite model, consisting of encoder and decoder parts, trained on Kinetics-400 dataset. The encoder model uses Video Transformer approach with ResNet34 encoder. Please refer to the kinetics dataset specification to see list of action that are recognised by this composite model.

Example

Composite model specification

Metric Value
Source framework PyTorch*

Encoder model specification

The action-recognition-0001-encoder model accepts video frame and produces embedding. Video frames should be sampled to cover ~1 second fragment (i.e. skip every second frame in 30 fps video).

Metric Value
GFlops 7.340
MParams 21.276

Inputs

  1. name: "0", shape: [1x3x224x224] - An input image in the format [BxCxHxW], where:

    • B - batch size
    • C - number of channels
    • H - image height
    • W - image width

    Expected color order is BGR.

Outputs

The model outputs a tensor with the shape [1x512x1x1], representing embedding of processed frame.

Decoder model specification

The action-recognition-0001-decoder model accepts stack of frame embeddings, computed by action-recognition-0001-encoder model.

Metric Value
GFlops 0.147
MParams 4.405

Inputs

  1. name: "0" , shape: [1x16x512] - An embedding image in the format [BxTxC], where:
    • B - batch size.
    • T - Duration of input clip.
    • C - dimension of embedding.

Outputs

The model outputs a tensor with the shape [bx400], each row is a logits vector of performed actions.

Legal Information

[*] Other names and brands may be claimed as the property of others.