详情
推薦
JAV_mika_kayama_v1
JAV Mika_Kayama

JAV Mika_Kayama

1.7K
62
10
#jav
#gravure
#哪裡的男人
#名人
#日文
#?星
#寫實
#celebrity,

Imitation model. Modelled after retired Japanese gravure model and AV star Mika Kayama (佳山三花).

OK, many of you probably won't know this one. Mika Kayama's debut in AV is fairly short-lived, having started as a gravure model in ~2002 under the alisa of Arisa Oda (小田有紗), and only joined the JAV industry for 4 years before retiring in 2011. Some of her work and appearance do leave an impression among fans, especially those that follow the early days of JAV company s1 very closely. I personally find her petite physique and square-ish facial structure within a mature JAV genre quite unique, hence this model.

~~~~~~~~~~~~~~~~~~~~~~~~~~
Recommended LoRA strength: 0.6 (0.5-1.0 are plausible range dependent on checkpoints; higher values changes the shape of the face and enlarges the chin..; use the high weights for Adetailer instead)

Main Prompt:

mikakayama_jav, ???????, brown eyes, jewelry, long hair

variations: black eyes, medium hair, tied hair, hair bun, earrings

facial variations: parted lips, lipstick, lip gloss, blush

*Use of facial prompts depend on checkpoints, use it if your checkpoint have a bias towards minimal makeup, most of the time you wont need any of these

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Checkpoints that work well for this LoRA:

1) Realisian/RealisticMix. Back to the almost-perfect duo. I dont even need to cherry-pick, most pictures are quite on-point.


2) PerfectWorld and UltimateHentai My only complaint is the more 2.5D style render. Nothing major.

3) Other Asian-based photorealistic models. Only tried those, doesnt sway too far. Technically even merged models like GrayColorSemiReal-mix and ClearRealism-mix works well too.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Note for achieving good resemblance:

This LoRA is easy when it comes to reproducing good resemblance. 2 Steps:

1. Generate the main prompt using a LoRA strength of 0.6~0.7, higher for full body modes.

2. Supplement that by enabling face Adetailer, and set the weight slightly higher to 0.8. e.g.
"face of mikakayama_jav, long hair, brown eyes, parted lips, lipstick, <lora:JAV_mika_kayama_v1_epoch_8:0.8>"

The makeup prompts are optional, but can help create a more "mature female" appearance, if you wish.

It is still better to use AD at baseline generation (512,768 etc) and then upscale it later.

Recommended approach: is to not use HiRes, but upscale it with img2img instead.

Use a denoise strength of 0.25 -0.3 (anything higher will overcorrect face during the process), and turn off Adetailer.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Issues:

BadHands mainly. Nothing major.

Post-production edits/Reproducibility:

Minor inpainting on some Badhands, and clothing.

Happy prompting. And do share [SFW] creations~

查看译文

評分與評論

-- /5
0 個評分

尚未收到足夠的評分或評論

R
ryk_civit
168
2.8K
與模型對話
公告
2024-04-01
发布模型
2024-04-01
更新模型資訊
模型详情
類型
LORA
发布時間
2024-04-01
基础模型
SD 1.5
版本介绍

v1. Stable release @ epoch #8. trained using 99 High Quality images.

许可范围
來源: civitai

1. 轉载模型僅供學習與交流分享,其版權及最終解释權归原作者。

2. 模型原作者如需認領模型,請通過官方渠道联系海藝AI工作人員進行認證。我們致力於保護每一位創作者的權益。 點擊去認領

創作许可范围
在線生圖
進行融合
允许下载
商業许可范围
生成圖片可出售或用於商業目的
允许模型轉售或融合后出售
QR Code
下載SeaArt App
在移動端继續你的AI創作之旅