File size: 4,132 Bytes
6b4fb8a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e43d7ea
 
 
 
6b4fb8a
 
 
 
 
 
 
 
 
 
34f02ca
 
6b4fb8a
 
 
 
 
 
 
 
 
 
 
 
 
 
e43d7ea
 
6b4fb8a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
34f02ca
6b4fb8a
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
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79

# Animetic Light

[<img src="./assets/images/main_image.webp" alt="main image" width="400">](./assets/images/main_image.webp)

[![README - Japanese](https://img.shields.io/badge/README-Japanese-blue.svg?logo=markdown)](README_ja.md)
[![README - English](https://img.shields.io/badge/README-English-blue.svg?logo=markdown)](README.md)

アニメスタイルの画像を対象とした、背景生成およびリライティングのための実験的モデルです。
FramePackの1フレーム推論に対応したLoRAです。

写真のリライティングには、IC-Light V2 を推奨します。

- [IC-Light V2 (Flux-based IC-Light models) · lllyasviel IC-Light · Discussion #98](https://github.com/lllyasviel/IC-Light/discussions/98)
- [IC-Light V2-Vary · lllyasviel IC-Light · Discussion #109](https://github.com/lllyasviel/IC-Light/discussions/109)

IC-Light V2-Varyは、huggingface spaceで使えますが、IC-Light V2はfal.aiなどのプラットフォームでAPI経由で使用できます。

- [fal-ai/iclight-v2](https://fal.ai/models/fal-ai/iclight-v2)

## 機能

- プロンプトに基づいて背景を生成し、キャラクター部分は保持したままリライティングを行います。

## 生成例

[<img src="./assets/images/sample.webp" alt="animatic light sample" width="500">](./assets/images/sample.webp)

## 使い方

- 画像解像度は総ピクセル数1024x1024がおすすめです。

- シンプルでフラットな背景を持つ画像をご用意ください。
または背景のある画像の場合、その背景部分を大まかに単色で塗りつぶすだけでも、機能することがあります。  

[<img src="./assets/images/erased_bg_sample.webp" alt="erase bg sample" width="300">](./assets/images/erased_bg_sample.webp)

- プロンプトは、Gemini Flash 2.5での作成を推奨します。
プロンプトの長さは700文字以上(スペース、改行含む)が推奨されます。
プロンプト生成に役立つGemini Flash 2.5用のプロンプトは[こちら](./prompt.txt)。  
画像と簡単な説明を入力すると、補完されたプロンプトが出力されます。
  - 例1: <ドラゴンの画像\> "ドラゴンが青い炎を吹きながら飛んでいる。背景は火山。"
  - 例2: <ドラゴンの画像\> "火山"
- プロンプトの先頭に以下のタグを付与します。  
`illustration style`, (オプション: `ambient lighting`, `dim lighting`)  

- 注意: 対応していない画像スタイル(フィギュア風)はノイズが出たり、背景が残る場合があります。

### ComfyUIワークフロー

ワークフローファイルは[こちら](./workflows/workflow_animetic_light.json)  
サンプル画像は[こちら](./sample_images/dragon.webp)  
[xhiroga/ComfyUI-FramePackWrapper_PlusOne](https://github.com/xhiroga/ComfyUI-FramePackWrapper_PlusOne)をご使用ください。  
モデルのインストール方法については、各推論環境のドキュメントをご参照ください。  
また、`animetic_light.safetensors``ComfyUI/models/loras`ディレクトリに配置してください。

## リソース

### FramePack

- [GitHub - lllyasviel/FramePack: Lets make video diffusion practical!](https://github.com/lllyasviel/FramePack)

### 推論環境

- [GitHub - xhiroga/ComfyUI-FramePackWrapper_PlusOne](https://github.com/xhiroga/ComfyUI-FramePackWrapper_PlusOne)
- [GitHub - kijai/ComfyUI-FramePackWrapper](https://github.com/kijai/ComfyUI-FramePackWrapper)
- [GitHub - git-ai-code/FramePack-eichi](https://github.com/git-ai-code/FramePack-eichi)

### 学習環境

- [GitHub - kohya-ss/musubi-tuner](https://github.com/kohya-ss/musubi-tuner)
  - [musubi-tuner/docs/framepack_1f.md at main · kohya-ss/musubi-tuner](https://github.com/kohya-ss/musubi-tuner/blob/main/docs/framepack_1f.md)

---

データセット作成・その他のお仕事募集中!メールアドレスかXのDMまでご連絡ください。  
<orcoki2000@gmail.com>  
[![x - ippanorc](https://img.shields.io/badge/ippanorc-7e913b.svg?logo=x)](https://x.com/ippanorc)