liveBook · Manning

Rank-5 tensor

As explained in Deep Learning with R, Third Edition

A rank-5 tensor is a five-axis array commonly used for batches of videos.

A rank-5 tensor stores video data as five axes: samples, frames, height, width, and channels. Each sample is a sequence of images. section 2.2.8

Video batches

A batch of videos has shape (samples, frames, height, width, channels). The samples axis selects individual videos; frames represents their sequence of images; height and width describe each image; channels store each pixel’s values. section 2.2.8

Read the five axes in this order:

How Deep Learning with R, Third Edition covers it

How deep it goes
Defines the five-axis shape and axis meanings, without explaining implementation or model architecture.
Examples
Uses batches of videos composed of ordered frames, each frame represented as an image.
What it assumes
Assumes familiarity with arrays and images while relating videos to simpler tensor categories.
What you can do after
Evaluate video data layouts by identifying samples, frames, spatial dimensions, and channels.
How it connects
Connects video tensors with images, sequence data, batches, and lower-rank tensor representations.
Cover of Deep Learning with R, Third Edition
liveBookmanning.com
Read Deep Learning with R, Third Edition on liveBook Get the book at manning.com

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Deep Learning with R, Third Edition. The section links point to the parts of the book it draws on.