[FEATURE] hub.auto for audio and video htypes

[FayazRahman]

🚨🚨 Feature Request

• Related to an existing Issue
• A new implementation (Improvement, Extension)

Is your feature request related to a problem?

Extend hub.auto to enable ingesting audio and video data. Take a look at ingesting image datasets for a start.

Description of the possible solution

Adding a video_ingestion.py file to hub/auto/unstructured with necessary changes would work. But this might mean rewriting a lot of code as in image_classification.py

An alternative solution to the problem can look like

To avoid this, a general class could be written and used as base class for image, audio and video.

[neel2299]

Can I work on this issue ??

[FayazRahman]

@neel2299 Of course. I'll assign it to you.

[neel2299]

Hello!! Can we add dependencies to get the video/audio meta data ?? @FayazRahman @farizrahman4u @mikayelh

[farizrahman4u]

Hello!! Can we add dependencies to get the video/audio meta data ?? @FayazRahman @farizrahman4u @mikayelh

No, hub already has all the dependencies required for reading meta data from audio and video.

[farizrahman4u]

Also note that we are switching from miniaudio to PyaAV for audio in this PR : #1576 , so it would be ideal if you base your PR against it.

[vaibhavjack11]

I am Interested to work on this issue

[neel2299]

Noted, Thanks for the heads up. I will check the PR out.

[aryatalathi]

I would like to contribute to this issue. Can you please elaborate?

[mikayelh]

hi @aryatalathi, do you prefer working on this issue or the datase tupload issue? I would recommend working on one issue at a time.

[aryatalathi]

Yeah, sure! I will prefer the dataset upload issue. Thank you!

[mikayelh]

gotcha, good luck - let us know if you have any questions and hit me up in community slack for some guidance.