Develop AI systems to efficiently learn new skills and solve open-ended problems, rather than depend exclusively on AI systems trained with extensive datasets
How this post is connected to other assets
The order in which we perform tasks in the algorithm certainly matters. We see here that we are expanding the bars towards the gray bar, but we need to start with the bar furthest from the gray to ensure that we get the overlay ordering right.
For example, in the first train/test example, yellow needs to go "behind" green. We can assume that would work because, even though we draw yellow where the green should fill, the green is drawn second so it will replace the yellow there.
There's certainly other ways would can think about this. Instead of ordering and replacement, we move to 3D and then there's no overlap, just overshadowing.
We'd need to learn the Z dimension from the output instead of just assuming some fixed mapping between color and Z.
Discover other posts like this one