Visual Anagrams Model with Different Animals' Pairs
IllusionDiffusion Model with Different Animals' Pairs
Here's the draft version (only containing one example)
Link hereClassic Duck-Rabbit Concept
First prompt: "a duck", second prompt: "a rabbit"
First prompt: "a rabbit", second prompt: "a duck"
Hyperparameters tests on two generative optical illusion diffusion models
First prompt: "a duck", second prompt: "a rabbit"
1. Different Seeds (inference_steps=500, guidance scale=7.5)
2. Different Inference Steps (seed=42, guidance scale=7.5)
3. Different Guidance Scales (inference_steps=500, seed=42)
First prompt: "a rabbit", second prompt: "a duck"
1. Different Seeds (inference_steps=500, guidance scale=7.5)
2. Different Inference Steps (seed=42, guidance scale=7.5)
3. Different Guidance Scales (inference_steps=500, seed=42)
Classic Duck-Rabbit Concept
First prompt: "a duck", second prompt: "a rabbit"
First prompt: "a rabbit", second prompt: "a duck"
Addtional Info
1.IllusionDiffusion (Backbone: Stable Diffusion V1.5); 2.Visual Anagrams (Backbone: DeepFloyd)
Here's the draft version (only containing one example)
Link here
Our survey also supports mobile access. You can scan the QR code to take the survey on your phone. :)