File size: 2,335 Bytes
12385f1
 
 
 
 
 
 
 
 
 
99b34c3
00442d8
12385f1
 
 
036a4e5
12385f1
00442d8
12385f1
00442d8
12385f1
99b34c3
 
 
12385f1
00442d8
12385f1
99b34c3
12385f1
00442d8
12385f1
fca8c3e
 
 
 
 
12385f1
99b34c3
12385f1
00442d8
 
 
 
 
 
 
fca8c3e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
---
license: cc-by-4.0
tags:
- super-resolution
pretty_name: DF2K_BHI
size_categories:
- 1K<n<10K
---
# DF2K_BHI SISR Dataset

This is a filtered version of the DF2K dataset, specifically aimed for single image super-resolution model training.  
It is a result, or is one of the testing sets, for [my huggingface community blog post on my BHI-Filtering method](https://huggingface.co/blog/Phips/bhi-filtering).  

## Dataset Details

This is based on the DF2K dataset, which consists of 3450 images.  

These images then have been tiled into 21'387 512x512px tiles for faster sisr I/O training speed.  

The BHI filtering has been applied with the following default threshold values:  

Blockiness < 30  
HyperIQA >= 0.2  
IC9600 >= 0.4  

Which results in this DF2K_BHI dataset with 12'639 512x512 training tiles.  

## Visualization

Here is a visual example of the first few tiles that are in the dataset:  

<figure>
  <img
  src="https://cdn-uploads.huggingface.co/production/uploads/634e9aa407e669188d3912f9/KL2FnhtUBSCeAdjh4ZnoZ.png">
  <figcaption>Visual example - first few tiles of the DF2K_BHI Dataset</figcaption>
</figure>

## Training

As can be seen in the following training validation tensorboard graphs (DIV2K validation dataset), this BHI-filtered version of the DF2K dataset is able to achieve higher or similiar metric scores while seeing a reduction of -40.9% in training tiles quantity.   

Here is the training validation metrics on the DIV2K validation dataset, using the plksr_tiny architecture option, in comparison to the tiled full DF2K dataset:  

Tiled DF2K dataset (here called X): 21'387 tiles  
DF2K_BHI dataset: 12'639 tiles  

<figure>
  <img
  src="https://cdn-uploads.huggingface.co/production/uploads/634e9aa407e669188d3912f9/z_wh3koK9bpkRKjoWm4LH.png">
  <figcaption>Tensorboard: SSIM validation scores of DF2K and DF2K_BHI</figcaption>
</figure>
<figure>
  <img
  src="https://cdn-uploads.huggingface.co/production/uploads/634e9aa407e669188d3912f9/6RkW2uILyHTNhWrPLJqGx.png">
  <figcaption>Tensorboard: PSNR validation scores of DF2K and DF2K_BHI</figcaption>
</figure>
<figure>
  <img
  src="https://cdn-uploads.huggingface.co/production/uploads/634e9aa407e669188d3912f9/Ff9RlvfZ-89jDvsDhzQNB.png">
  <figcaption>Tensorboard: DISTS validation scores of DF2K and DF2K_BHI</figcaption>
</figure>