Upload 19 files
Browse files- cc-by-nc-sa/Alsa/4x_realistic_misc_alsa.pth +3 -0
- cc-by-nc-sa/EzoGaming/2x_KemonoScale_v2.pth +3 -0
- cc-by-nc-sa/Foolhardy/4x_foolhardy_Remacri.pth +3 -0
- cc-by-nc-sa/Kim2091/4x-UltraSharp.pth +3 -0
- cc-by-nc-sa/Kim2091/4x-UniScale-Balanced [72000g].pth +3 -0
- cc-by-nc-sa/Kim2091/4x-UniScale-Interp.pth +3 -0
- cc-by-nc-sa/Kim2091/4x-UniScale-Strong [42400g].pth +3 -0
- cc-by-nc-sa/Kim2091/4x-UniScaleNR-Balanced [34400g].pth +3 -0
- cc-by-nc-sa/Kim2091/4x-UniScaleNR-Strong [62400g].pth +3 -0
- cc-by-nc-sa/Kim2091/4x-UniScaleV2_Mod.Shapr.Soft.INFO.txt +31 -0
- cc-by-nc-sa/Kim2091/4x-UniScaleV2_Moderate.pth +3 -0
- cc-by-nc-sa/Kim2091/4x-UniScaleV2_Sharp.pth +3 -0
- cc-by-nc-sa/Kim2091/4x-UniScaleV2_Soft.pth +3 -0
- cc-by-nc-sa/Kim2091/4x-UniScale_Restore.pth +3 -0
- cc-by-nc-sa/Kim2091/4x-UniScale_Restore_INFO.txt +1 -0
- cc-by-nc-sa/Kim2091/LICENSE.txt +1 -0
- cc-by-nc-sa/LyonHrt/atrib.txt +1 -0
cc-by-nc-sa/Alsa/4x_realistic_misc_alsa.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3f379331b50e54efea46a3221ecb107b1ff369bba65f3f0a9468b29f815406bb
|
3 |
+
size 66960011
|
cc-by-nc-sa/EzoGaming/2x_KemonoScale_v2.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b64cec8f4c2a848e0e67eb8d8992b8dd9e0d6a82b330f09150c36b3089d9878b
|
3 |
+
size 66758126
|
cc-by-nc-sa/Foolhardy/4x_foolhardy_Remacri.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e1a73bd89c2da1ae494774746398689048b5a892bd9653e146713f9df8bca86a
|
3 |
+
size 67025055
|
cc-by-nc-sa/Kim2091/4x-UltraSharp.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a5812231fc936b42af08a5edba784195495d303d5b3248c24489ef0c4021fe01
|
3 |
+
size 66961958
|
cc-by-nc-sa/Kim2091/4x-UniScale-Balanced [72000g].pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:abfd3fbcec247c88537b20acca1f023e7a11b7162e498145295a8461e291dcd2
|
3 |
+
size 66961958
|
cc-by-nc-sa/Kim2091/4x-UniScale-Interp.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:dda528d5ecf32b250d0e2ad904c71ebf297f51a5dc44b32cca618d8fe35a8919
|
3 |
+
size 67010245
|
cc-by-nc-sa/Kim2091/4x-UniScale-Strong [42400g].pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5d80a6b236e8ee9632bb4df6a319635d0f0cb2889b368a884b6fcc914f12bbee
|
3 |
+
size 66961958
|
cc-by-nc-sa/Kim2091/4x-UniScaleNR-Balanced [34400g].pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8377f8b71b792bbe298f86e673f8b2d19d083f9143880fbbb51c6a569b1c90c9
|
3 |
+
size 66961958
|
cc-by-nc-sa/Kim2091/4x-UniScaleNR-Strong [62400g].pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d5d3200fe2909c63384af88dd6b35d5572da3035f6f96ad9a3e6dd5e05480645
|
3 |
+
size 66961958
|
cc-by-nc-sa/Kim2091/4x-UniScaleV2_Mod.Shapr.Soft.INFO.txt
ADDED
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
@News @Wiki Editor
|
2 |
+
**Name:** 4x-UniScaleV2_Soft/Moderate/Sharp
|
3 |
+
**Author:** Kim2091
|
4 |
+
**License:** CC BY-NC-SA 4.0
|
5 |
+
**Link:** <https://mega.nz/folder/zZhA1KoD#ds2nmgDNV4hfFpNSfEqN6Q>
|
6 |
+
**Model Architecture:** ESRGAN
|
7 |
+
**Scale:** 4
|
8 |
+
**Purpose:** Effectively a UniScale successor. It does nearly everything better, other than dealing with noise or compression. Use UniScale_Restore or UniScale_Iterp for that.
|
9 |
+
|
10 |
+
**Iterations:** 111k
|
11 |
+
**batch_size:** 4
|
12 |
+
**HR_size:** 112
|
13 |
+
**Epoch:** 8
|
14 |
+
**Dataset:** Custom Dataset consisting of lossless 4k frames from Metal Arms: Glitch in the System, Just Cause 3, Dirt 3, Forza Horizon 3, Sleeping Dogs, and self-edited photos from SignatureEdits + ATLA DVD images
|
15 |
+
**Dataset_size:** 18,909 tiles + 288 ATLA Frames
|
16 |
+
**OTF Training** Yes (JPEG artifacts, base_blur, bsrgan_resize)
|
17 |
+
**Pretrained_Model_G:** 4xESRGAN
|
18 |
+
|
19 |
+
**Description:** I really don't have a great description for this model set. It just works well on nearly everything (other than Anime sadly). They work best with realistic images. I hope you guys like it :stuck_out_tongue:
|
20 |
+
|
21 |
+
The Moderate and Soft models are interpolations between UniScaleV2_Sharp, UniScale-Strong, and UniScaleNR-Balanced.
|
22 |
+
|
23 |
+
If your image has compression, the Soft model will work the best. Moderate and Sharp will work on images with compression, but they won't be as clean.
|
24 |
+
|
25 |
+
UniScale_Restore or UniScale_Interp (available in the main UniScale folder) will likely work better for images with heavy compression.
|
26 |
+
|
27 |
+
Comparisons:
|
28 |
+
https://cdn.discordapp.com/attachments/884239326471393331/884296266857713704/unknown.png
|
29 |
+
https://cdn.discordapp.com/attachments/884239326471393331/884294468709257216/unknown.png
|
30 |
+
https://cdn.discordapp.com/attachments/884239326471393331/884298572894441512/unknown.png
|
31 |
+
https://cdn.discordapp.com/attachments/884239326471393331/884302370136289323/unknown.png
|
cc-by-nc-sa/Kim2091/4x-UniScaleV2_Moderate.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:229eddb4e6d3546bc75ac5b19eeced11522ecaf57be9da227d541edf9435605d
|
3 |
+
size 67010245
|
cc-by-nc-sa/Kim2091/4x-UniScaleV2_Sharp.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f364528f5c9b0bc0543dfcafc8991de95a274b1b54ad5769313e9455a50c48ee
|
3 |
+
size 66961958
|
cc-by-nc-sa/Kim2091/4x-UniScaleV2_Soft.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5e2b1d5272d22728e6c14251c1197ae58872650a008a9f2c887b9c394c20e4f4
|
3 |
+
size 67010245
|
cc-by-nc-sa/Kim2091/4x-UniScale_Restore.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7309d334680427f433266cc14ae5a64d9c9be2ab0e7a4d887ce561b0aa4f5f1a
|
3 |
+
size 66961958
|
cc-by-nc-sa/Kim2091/4x-UniScale_Restore_INFO.txt
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
This model has strong compression removal that helps with restoring heavily compressed or noisy images. It is intended to compete with BSRGAN.
|
cc-by-nc-sa/Kim2091/LICENSE.txt
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
These models are licensed under CC BY-SA 4.0
|
cc-by-nc-sa/LyonHrt/atrib.txt
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
https://upscale.wiki/wiki/User:LyonHrt
|