Hugging Face's logo

zelaki
/
eq-vae-ldm

Model card Files Community
zelaki commited on
Commit
44bee05
·
verified ·
1 Parent(s): c30156b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +8 -5
README.md CHANGED
@@ -1,12 +1,14 @@
1
 
 
2
  ## EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling
3
  Arxiv: https://arxiv.org/abs/2502.09509 <br>
4
 
 
5
  **EQ-VAE** regularizes the latent space of pretrained autoencoders by enforcing equivariance under scaling and rotation transformations.
6
 
7
  ---
8
  #### Model Description
9
- This model is a regularized version of [SD-VAE](https://github.com/CompVis/latent-diffusion). We finetune it with EQ-VAE regularization for 44 epochs on Imagenet with EMA weights.
10
 
11
 
12
  ## Model Usage
@@ -19,9 +21,10 @@ Reconstruction performance of eq-vae-ema on Imagenet Validation Set.
19
 
20
  | **Metric** | **Score** |
21
  |------------|-----------|
22
- | **FID** | 0.552 |
23
- | **PSNR** | 26.158 |
24
- | **LPIPS** | 0.133 |
25
- | **SSIM** | 0.725 |
26
  ---
27
 
 
 
1
 
2
+
3
  ## EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling
4
  Arxiv: https://arxiv.org/abs/2502.09509 <br>
5
 
6
+
7
  **EQ-VAE** regularizes the latent space of pretrained autoencoders by enforcing equivariance under scaling and rotation transformations.
8
 
9
  ---
10
  #### Model Description
11
+ This model is a regularized version of [SD-VAE](https://github.com/CompVis/latent-diffusion). We finetune it with EQ-VAE regularization for 5 epochs on OpenImages.
12
 
13
 
14
  ## Model Usage
 
21
 
22
  | **Metric** | **Score** |
23
  |------------|-----------|
24
+ | **FID** | 0.82 |
25
+ | **PSNR** | 25.95 |
26
+ | **LPIPS** | 0.141 |
27
+ | **SSIM** | 0.72 |
28
  ---
29
 
30
+