Confusing Coleopterists
AI
generative image model
SHOW HN2020
Legend
Confusing Coleopterists
A beetle generator made by machine-learning zoological illustrations
cunicode.com#stylegan#beetle generator#gan#machine learning illustrations
How HN took it
3 positive, 13 neutral or asking, 0 mixed, 3 critical, out of 19 comments about the project. Thread score +0.02.
Idea: 2 for, 1 againstExecution: 0 for, 2 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Clever and beautiful results, and fun to browse
From 3 comments
This is a beautiful write-up. And I don't wish to detract from the author's work. Nature's result side-by-side with that of the Machine. Makes one feel as though we've taken a…
ArtWomb on HNI wonder if this could be used as an identification tool for beetles like a phantom sketch. You could move some sliders to get it closer to the bug you are thinking of or which…
bayesian_horse on HNExcellent. I did something similar a few months ago using a dataset of zoological silhouettes, resulting in a menagerie of mammals, bugs, spiders and other mutant wonders.…
cortesi on HN
Objections
Mirrored halves look unnatural because lighting is asymmetric
From 1 comment
Most of the images have lighting which isn’t symmetrical and I think that makes it pretty obvious when an image is made by mirroring half a beetle. But maybe there’s some way to…
dan-robertson on HN
Legs and symmetry are handled poorly
From 1 comment
Interesting to see how it handled the legs. Looks like it had problems with the fact insects have 3 pairs of symmetrical legs.
proc0 on HN
Requests
Check the model for overfitting against source images
From 2 comments
Was this model tested for overfitting? I do not have any sense of whether the beetles that I'm seeing match some source pictures exactly. I noticed that the transformations seem…
fiter on HNIs there a good way to test a gan for overfitting?
enchiridion on HN
Train on Chernobyl mutated insect images
From 1 comment
Would be interesting to see the result trained on images from this book[1] about mutation on insects in the chernobyl area.…
eskimobloood on HN
Questions
Is the model available to run?
From 1 comment
Is the source/model/network available to run?
jamesfisher on HN
Does the network draw the beetles or how does it generate pixels?
From 2 comments
What I never understand about these things is ... what actually does the drawing? The AI decides what the beetle looks like, at what level of abstraction? When/how does it go…
nsxwolf on HNVery clever. Do you know if a machine does the drawing? Wasn't totally clear on that bit.
slynn12 on HN
Alternatives named
Picture This plant identification app
From 1 comment
They already have similar apps to this for a range of things. I guess the most popular ones are for plants [1] [1] https://www.picturethisai.com/
deanclatworthy on HN
Dawkins biomorphs
From 1 comment
Reminds me of Dawkins' biomorphs: http://www.emergentmind.com/biomorphs
auton1 on HN