Sign In

Stereoscopic Portrait - 3D 立体肖像

34
551
651
6
Verified:
SafeTensor
Type
LoRA
Stats
392
585
Reviews
Published
Sep 27, 2023
Base Model
SD 1.5
Training
Steps: 2,004
Epochs: 12
Hash
AutoV2
4F44FD132C
default creator card background decoration
Lunar Love Contest Participant
LordJia's Avatar
LordJia

This is an experimental project designed to produce Stereoscopic Portraits. It supports both Parallel-view and Cross-view modes, corresponding to two downloadable versions.

Performance:

When used as recommended, the subject of the portrait achieves a relatively high success rate. The background, in contrast, might require multiple attempts and might achieve a lower success rate.

User Guide: For optimal results and image quality, strictly follow the procedure outlined below:

  • Checkpoint: Realistic Vision V5.1

  • Textual Inversion: ng_deepnegative_v1_75t

  • Positive prompts: {prompt}, masterpiece, high detail, 8k, high detailed skin, 8k uhd, high quality

  • Negative prompts: ugly, disfigured, deformed, worst quality, low quality, ng_deepnegative_v1_75t

  • LoRA Weight: 0.8-1.0

  • Image Dimensions: Width: 832, Height: 560

  • Optional: Use Hires. fix to enhance image quality with a Denoising strength of 0.3.

Note:

  • Using the Realistic Vision V5.1 in tandem with ng_deepnegative_v1_75t can significantly improve the success rate and generated image quality. See the "Suggested Resources" section below for more details.

  • This LoRA is specifically designed for generating Stereoscopic Portraits with a human subject. Without a human subject, a 3D effect cannot be achieved.

  • For detailed information on Stereoscopy, refer to the Wikipedia entry titled "Stereoscopy".


这是一个实验性项目,用来生成 立体视觉肖像(Stereoscopic Portrait),支持 Parallel-viewCross-view 两种模式(对应两个下载版本)。

在按照推荐方法使用的前提下,人物主体可以获得比较高的成功率,背景的成功率相较人物主体要低一些,有时需要多次尝试。

使用说明:(严格按照以下流程生成图片,可以获得比较高的成功率与图像质量。)

  • 模型(checkpoint):Realistic Vision V5.1

  • 嵌入式(Textual Inversion):ng_deepnegative_v1_75t

  • 正面提示词:{prompt}, masterpiece, high detail, 8k, high detailed skin, 8k uhd, high quality

  • 负面提示词:ugly, disfigured, deformed, worst quality, low quality, ng_deepnegative_v1_75t

  • LoRA 权重:0.8-1.0

  • 图像尺寸:宽度:832,高度:560

  • 可选:使用高清修复(Hires. fix)提升图像质量,重绘幅度(Denoising strength)0.3

注意:

  • 配合使用 Realistic Vision V5.1 与 ng_deepnegative_v1_75t 可以极大提高成功率与生成图像质量。具体见下方“Suggested Resources”。

  • 此 LoRA 只用于生成带有人物主体的立体视觉肖像图,如果不带有人物主体则无法产生立体效果。

  • 关于立体视觉(Stereoscopy)的具体说明,参见维基百科“立体图”词条。