Wouldn't that be complete overkill, though? Zooming and panning are so much simpler and less compute intensive.
Not having a bunch of photos of people pointing and just zooming/panning to that bit.
Labelling itself already solves the issue (i.e. a bunch of photos with people pointing in certain directions).
So basically you'd solve the problem you want to solve in order to automate a task that you already performed.
Not necessarily.
One could train a model on synthesized data using e.g. blender, some programmable 3D people models with hand controls and some generic background images to paste them on.
It's a neat idea, and I hope it becomes more popular.
Me specifically or in general?
Anyway, for a Multimedia Information Retrieval course I chose to do my term project on training a neural network with synthetic data. In particular I modded Minecraft such that when I press a button it saves two screenshots: one regular and one where the game renders a depth map instead. I used this to generate ~1000 samples with perfectly accurate depth maps. Because of the mods, texture pack and world I used the data was somewhat realistic: https://i.stack.imgur.com/Zai51.jpg https://i.stack.imgur.com/eamMR.png
This data was then used to train a neural network to predict the depth map of unseen images. It was relatively successful, but requires more data and more research, I only had so much time for a term paper.
Bonus points for candid photos, no ads, low resolution images and nostalgic scenes.
I like this structure of HN. Kinda like old Slashdot. Something funny and lighthearted amongst all the heavy articles.