Supporting video inputs (and even outputs) would be a very nice feature to have.
The main issue is how do we reduce something as huge as a video to a representation that's e.g. < 10,000 float32's long.
This is not really a priority of the Mindsdb team right now, since we don't have a usecase for it, but if anyone does and would like to help we'd be very glad to accept external contributions
DATA-X: m420 - Bread & Butter Deep Learning: Regression and Classification using TensorFlow v2 and Ludwig. This notebook covers advanced topics in machine learning. However, it does not require any prior knowledge in machine learning. The goal of this notebook is to teach a user how to deploy deep learning regression and classification models, using structured data. This is task is so common to machine learning, that it is pretty much the bread and butter of ML engineers.
Supporting video inputs (and even outputs) would be a very nice feature to have.
The main issue is how do we reduce something as huge as a video to a representation that's e.g. < 10,000 float32's long.
This is not really a priority of the Mindsdb team right now, since we don't have a usecase for it, but if anyone does and would like to help we'd be very glad to accept external contributions