Update README.md
Browse files
README.md
CHANGED
|
@@ -13,7 +13,36 @@ The colab implementation can be found [here](https://colab.research.google.com/g
|
|
| 13 |
| -------- | --------- |
|
| 14 |
|  |  |
|
| 15 |
|
| 16 |
-
#
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
| -------- | --------- |
|
| 14 |
|  |  |
|
| 15 |
|
| 16 |
+
# Overview
|
| 17 |
+
The model is able to produce a wide variety of images. This is likely owing to the use of [minibatch discrimination](https://arxiv.org/abs/1606.03498v1).
|
| 18 |
+
|
| 19 |
+
The quality of the generated images is reminiscent of the early days of GANs.
|
| 20 |
+
|
| 21 |
+
## Left-facing planes
|
| 22 |
+
|
| 23 |
+
|
| 24 |
+
## Right-facing planes
|
| 25 |
+
|
| 26 |
+
|
| 27 |
+
## Planes on the ground
|
| 28 |
+
|
| 29 |
+
|
| 30 |
+
## Poor quality images
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
## Unique generations
|
| 34 |
+
|
| 35 |
+
|
| 36 |
+
# Image memorization
|
| 37 |
+
|
| 38 |
+
To check whether the Generator is simply memorizing particular images, latent space exploration can be used.
|
| 39 |
+
|
| 40 |
+
The latent vector for a generated image is slightly mutated, and the new images are compared to the original. The comparisons (shown below) suggest that mutations in the latent space correspond to small (sometimes imperceptible) mutations in the generated images.
|
| 41 |
+
|
| 42 |
+
These observations support the idea that the generator's space is continuous, and the generator is not restricted to a finite number of modes.
|
| 43 |
+
|
| 44 |
+
| Original Generated Image | Neighbors in latent space | Diff against original image |
|
| 45 |
+
| ----------- | -------- | ----------- |
|
| 46 |
+
|  |  |  |
|
| 47 |
+
|  |  |  |
|
| 48 |
+
|  |  |  |
|