Upscale Models
License: CC-BY-SA-4.0
Link: Download Github
Model Architecture: RealPLKSR
Scale: 2
Purpose: Upscale text in very low quality to normal quality.
Iterations: 147k
batch_size: 8
patch_size: 48
Dataset: 8'000 images
Pretrained_Model_G: No
Description: The upscale model is specifically designed to enhance lower-quality text images, improving their clarity and readability by upscaling them by 2x. It excels at processing moderately sized text, effectively transforming it into high-quality, legible scans. However, the model may encounter challenges when dealing with very small text, as its performance is optimized for text of a certain minimum size. For best results, input images should contain text that is not excessively small.
License: CC-BY-SA-4.0
Link: Download Github
Model Architecture: RealPLSKR
Scale: 4
Purpose: Proficient in upscaling high-resolution photos to medium quality, preferably with a minimum size of 300px on the smallest side.
Iterations: 152k
batch_size: 8
HR_size: 1024
Dataset: 8684
OTF Training: No
Pretrained_Model_G: No
Description: Skilled in working with cats, hair, parties, and creating clear images. Also proficient in resizing photos and enlarging large, sharp images. Can effectively improve images from small sizes as well (300px at smallest on one side, depending on the subject). Experienced in experimenting with techniques like upscaling with this model twice and then reducing it by 50% to enhance details, especially in features like hair or animals.
Upscale example:
License: CC-BY-SA-4.0
Link: Download Github
Model Architecture: RealPLKSR
Scale: 2
Purpose: Upscale VHS and restore image quality.
Iterations: 385k
batch_size: 8
patch_size: 48
Dataset: 6'019 images
Pretrained_Model_G: No
Description: An advanced VHS recording model designed to enhance video quality by reducing artifacts such as haloing, ghosting, and noise patterns. Optimized primarily for PAL resolution (NTSC might work good as well).
License: CC-BY-SA-4.0
Link: Download Github
Model Architecture: compact
Scale: 4
Purpose: Skilled in upscaling anime content sourced from DVDs to higher resolutions.
Iterations: 328k
batch_size: 4
HR_size: 192
Dataset: 6358 handpicked images
OTF Training: No
Pretrained_Model_G: No
Description: Capable of upscaling anime content from DVD sources by a factor of four, resulting in slightly softened lines while preserving the original colors. Works perfect with 90's cartoon series.
License: CC-BY-SA-4.0
Link: Download Github
Model Architecture: span
Scale: 4
Purpose: Proficient in upscaling high-resolution photos to medium quality, preferably with a minimum size of 300px on the smallest side.
Iterations: 488k
batch_size: 4
HR_size: 256
Dataset: 6500
OTF Training: No
Pretrained_Model_G: No
Description: Skilled in working with cats, hair, parties, and creating clear images. Also proficient in resizing photos and enlarging large, sharp images. Can effectively improve images from small sizes as well (300px at smallest on one side, depending on the subject). Experienced in experimenting with techniques like upscaling with this model twice and then reducing it by 50% to enhance details, especially in features like hair or animals.
Upscale example:
License: CC-BY-SA-4.0
Link: Download Github
Model Architecture: compact
Scale: 4
Purpose: Fast upscale with high-resolution photos to medium quality, preferably with a minimum size of 400px on the smallest side.
Iterations: 499k
batch_size: 4
HR_size: 256
Dataset: 6500
OTF Training: No
Pretrained_Model_G: No
Description: Skilled in working with cats, hair, parties, and creating clear images. Also proficient in resizing photos and enlarging large, sharp images. Can effectively improve images from small sizes as well (300px at smallest on one side, depending on the subject). Experienced in experimenting with techniques like upscaling with this model twice and then reducing it by 50% to enhance details, especially in features like hair or animals. This model is however not as good as the span version of this model, but close enough.
Upscale example:
