What Science Can Do With Computer Vision Using Cutting-Edge XAI
A Talk by Phil Teare
Director, Computer Vision,
About this talk
Deep Learning is frequently referred to as a ‘Black Box’ method.
While the inputs and outputs are known, the complexity of what happens within them makes understanding and communicating how they make their predictions and estimates intractable.
If the hidden, or latent layers of processing are crucial to how a neural network produces its output, how might we better understand the mechanisms at play, and what becomes possible when we can achieve this?
In this talk, we will look at a process referred to at AstraZeneca, as Direct Counterfactual Rendering. What it is capable of, how it differs from other explainability techniques and when we lift the lid on the black box, what insights come pouring out?
Stages covered by this talk
Proudly supported by
Want to sponsor this event? Contact Us.