kontext lora | Remove everything_scene object removal
Back

kontext lora | Remove everything_scene object removal
164 2 34

Accentuation

Composition

ACG design

kontext lora | Remove everything_scene object removal

Kontext_lora_Everything Removal (Remove/Erase Items in the Scene Enhancement)

Everything Removal lora is based on the kontext workflow for image processing and cannot be used independently. By incorporating this lora into the workflow and following the basic format of prompts (commands), the removal of items in scenes/background images can be greatly improved. The original version requires highly specific prompts, and even then, the removal may not be stable. Adding this lora significantly enhances the stability and precision of removing items in scene images.


The standard format for prompts is as follows:

1. For simple items with high recognition rate: Remove the item in English

Example: Remove perfume bottle.  (移除香水瓶)

2. For complex items that AI might not recognize: Remove item name, color/shape, item location

Example: Remove blank label from the white tube (从白色管上移除空白标签)


Model weight: Set to 1

• Recommended step: 20 30

• Recommended sampler: Euler simple

• CFG: 2.5

遥远的炼丹师

遥远的炼丹师

Accentuation

Composition

ACG design

Model Information

Active
Model Type:
LoRA
Basic Model:
F1基础-Kontext
Resource Name:
models/loras/kontext-lora-万物消除V1.1_遥远的炼丹师.safetensors
MD5:
7fafc2aa53f883095c839c4b17ed2a6f

Kontext_lora_Everything Removal (Remove/Erase Items in the Scene Enhancement)

Everything Removal lora is based on the kontext workflow for image processing and cannot be used independently. By incorporating this lora into the workflow and following the basic format of prompts (commands), the removal of items in scenes/background images can be greatly improved. The original version requires highly specific prompts, and even then, the removal may not be stable. Adding this lora significantly enhances the stability and precision of removing items in scene images.


The standard format for prompts is as follows:

1. For simple items with high recognition rate: Remove the item in English

Example: Remove perfume bottle.  (移除香水瓶)

2. For complex items that AI might not recognize: Remove item name, color/shape, item location

Example: Remove blank label from the white tube (从白色管上移除空白标签)


Model weight: Set to 1

• Recommended step: 20 30

• Recommended sampler: Euler simple

• CFG: 2.5