owed art 1/5 ! character belongs to https://toyhou.se/fuzzybirb
actually very proud of this, i tried messing around with layers and stuff, and i think it turned out well ! :D
seen from Spain
seen from Türkiye
seen from United States
seen from China

seen from Indonesia

seen from Germany
seen from Iraq
seen from United Kingdom
seen from United States
seen from China
seen from United States
seen from Chile
seen from United States
seen from United States
seen from United Kingdom
seen from Singapore
seen from United States
seen from Russia
seen from Japan
seen from United Kingdom
owed art 1/5 ! character belongs to https://toyhou.se/fuzzybirb
actually very proud of this, i tried messing around with layers and stuff, and i think it turned out well ! :D
Air Over Layer | Sigma-HSE
The air over layer test is conducted to establish the onset temperature of exothermic activity of thin layers of material when exposed to elevated atmospheric temperatures. https://sigma-hse.com/thermal-stability-and-gas-and-vapour-testing/air-over-layer
The air over layer test is conducted to establish the onset temperature of exothermic activity of thin layers of material when exposed to elevated atmospheric temperatures. https://sigma-hse.com/thermal-stability-and-gas-and-vapour-testing/air-over-layer
Air Over Layer | Sigma-HSE
The air over layer test is conducted to establish the onset temperature of exothermic activity of thin layers of material when exposed to elevated atmospheric temperatures. https://sigma-hse.com/thermal-stability-and-gas-and-vapour-testing/air-over-layer
Testing Layers of Google’s DeepDreams
I’ve been testing out a bunch of the layers available as part of Google’s recently-released DeepDream code. Since tumblr will only allow so many photos at a time, here’s a summary of #layer test, all of which was completed with default settings:
Layer 1
All Layer 1: Very simplified, tends to come out looking like an overused Photoshop filter. Not my favorites.
Layer 2
conv2 and pool2: Still very simple, but a little less auto-filter looking. Starting to get interesting.
Layer 3
inception_3a and pool3: Probably my favorites for this underwater scene. Interesting detail without getting into dogslug territory.
inception_3b: Very painting-like. Still dogslug-free.
Layer 4
inception_4a: Starting to see eyeballs, but it’s not overly intrusive yet.
inception_4b: Gettin’ a little sluggy up in here.
inception_4c: With the exception of 1x1, getting a lot more eyes and new creatures.
inception_4d: The creatures are evolving.
inception_4e and pool4: Kind of a mess. Getting back into Photoshop auto-filter territory, but with extra legs.
Layer 5
inception_5a: Venturing into a totally new ecosystem. This would probably have the best effect with a higher number of steps/iterations.
inception_5b and pool5: Might work better on higher contrast, but in general not as exciting as some of the previous layers. Bumping up steps might help on this one too.
Conclusion: For this type of photo, Layer 3 appears to be the sweet spot where you get interesting images without being overwhelmed by genetic mutations. Changing the other settings may yield better results in the more complex layers.