Model Collection
					Collection
				
				9 items
				โข 
				Updated
					
				
Repository: https://huggingface.co/svjack/Kontext_Gray_Anime_Character_Colorize_lora
This repository provides a LoRA (Low-Rank Adaptation) fine-tuned variant of the Kontext image consistency model, specialized in colorizing grayscale anime character art while preserving facial features, skin tones, hair details. The model uses color reference images to achieve high-fidelity, context-aware colorization.
Precise Feature Retention
LoRA Efficiency
Prompt
| Grayscale Input | Reference Image | Colorized Output | 
|---|---|---|
|  |  |  | 
| Characteristics Preserved: Straw hat texture, scar under eye, muscle shading | 
| Grayscale Input | Reference Image | Colorized Output | 
|---|---|---|
|  |  |  | 
| Characteristics Preserved: Straw hat texture, scar under eye, muscle shading | 
| Grayscale Input | Reference Image | Colorized Output | 
|---|---|---|
|  |  |  | 
| Characteristics Preserved: Dodoco eyes, backpack details, hair ribbons | 
| Grayscale Input | Reference Image | Colorized Output | 
|---|---|---|
|  |  |  | 
| Characteristics Preserved: Feather accessories, armor gradients | 
| Grayscale Input | Reference Image | Colorized Output | 
|---|---|---|
|  |  |  | 
| Characteristics Preserved: Wing feather patterns, beak gradient | 
Explore more examples:
my_first_flux_kontext_lora_v1_000008500.safetensors