JAV Riho_Fujimori (藤森 里穂)

Imitation model. Modelled after Japanese AV star, TV personality and actress Riho Fujimori (藤森 里穂).

I realise I tend to do more actresses with long hair and huge 0ppai's. Well, apart from like 20 huge 0ppai JAV actresses, I do quite like some with slightly more modest 0ppai sizes. Riho is one of them.

Riho is considered one of the more recent favourites, stellar performance and broad media exposure (I personally rank her probably not too far off from Yua Mikami along popularity - totalling close to million followers on SocMed). She won several notable JAV awards like the [Valuable Actress of AV King] when she debuted under AV company SKY Perfect, and signed contracts with since late 2022. Recently, went into radio broadcasting (mini podcasts) earlier this month. Her physique is quite top-form, good amount of muscle and abs, cheerful vibe and plays just well into any JAV genre.

Now let's get to business.

~~~~~~~~~~~~~~~~~~~~~~~~~~
Recommended LoRA strength: 0.8 (0.6-1.0 are plausible range dependent on checkpoints; higher values >1 dont do much, apart from some background-bleeding; fairly stable LoRA in this case but dont go overboard with this..)

Main Prompt:

rihofujimori_jav, brown hair, brown eyes, blunt bangs, bangs, parted lips,variations: single hair bun, high ponytail, medium hair,

variations: parted lips, lipstick, makeup, blush, teeth, smile

Use of prompts depend on checkpoints, I didnt actually bother, since the main prompts work almost perfectly.

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

Checkpoints that work well for this LoRA:

1) Realisian. Like the name "realistic" "asian", almost gravure-like images. Not kidding.

3) RealisticMix. Has comparable performance with Realisian in resemblance. But my major issue with this is the exaggerated size. In retrospect, maybe I can try and control this in negative prompts(?)

2) PerfectWorld. Just too 2.5D to my liking. Still better than realistic European checkpoints (for this LoRA) though.

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

Note for achieving good resemblance:

Very straightforward to use. If you are lazy, you can just use the main prompt to control for everything and apply Adetailer to fix the faces.

Else, you can separate the prompts into 2 stages.

1. Generate the main prompt using a LoRA strength of 0.6-0.8. I did not run into any weird body proportion issues with 0.8. so its probably safe to say that this weight is stable for use.

2. Now depending on what you use for (1), turn on Adetailer, and put something like

"face of rihofujimori_jav, lips, short hair, brown hair, brown eyes, bangs, parted lips, realistic face, <lora:JAV_riho_fujimori_v1_epoch_10:0.8>"

Do remember that if you want her with her ponytail, make sure you set the prompts on both ends to reflect that.

set Adetailer denoise strength to 0.40-0.45. for Full body, you can try to set this higher to 0.50.

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 0.20 -0.25 (anything higher will overcorrect face during the process), and turn off Adetailer. I prefer 0.2 since I use foolhardy-reMarci.

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

Issues:

Nothing too noticeable, apart from checkpoints getting too excited with "". Easy problem to fix with negative prompts and re-seeds.

Post-production edits/Reproducibility:

Some badHand inpaintings. I got lazy and did not bother updating the prompt when running the batches of img2img upscale. So if you want to reproduce the images, let me know in comments. (actually, I just switch up the clothing prompts, everything else is the same - though you might need the right seed)

Happy prompting. And do share [SFW] creations~