Anime OTK

LoRA-DreamBooth

Sat, 25 Mar 2023 09:52:15 +0000

by: qqq123456

Description}: If you have any more questions about using LoRA, please feel free to leave a message below. I am trying to learn more about AI-related knowledge in spanking, and although it hasn't been smooth sailing, perhaps I will be able to produce more than just images of spankees in the near future.

3 comments

LoRA-DreamBooth

Sat, 25 Mar 2023 09:31:55 +0000

by: qqq123456

3 comments

LoRA !

Sat, 04 Mar 2023 02:27:10 +0000

by: qqq123456

Description}: Another picture of Makima was created using LoRA to eliminate the need for post-production in Photoshop. Despite the imperfect result of the drawing, there has been significant progress compared to previous attempts.

LoRA !

Sat, 04 Mar 2023 02:20:28 +0000

by: qqq123456

Description}: The LoRA model can be quickly mixed with the old model (to me this is more like a filter), and different LoRA can be mixed, which brings unexpected possibilities and convenience. The left image uses a single LoRA: makimaChainsawMan_offset The image on the right uses two LoRA: makimaChainsawMan_offset, spankingAss_v10 Note: The above pictures are not post-processed with Photoshop

Straddling Chair

Thu, 02 Mar 2023 15:32:22 +0000

by: qqq123456

Hands on table

Thu, 02 Mar 2023 15:30:19 +0000

by: qqq123456

Description}: AI still doesn't know much about the anatomy of a human hand or foot...but I think many models have improved on this recently.

2 comments

Embarrassing pose 2

Wed, 01 Mar 2023 16:39:36 +0000

by: qqq123456

Embarrassing poses

Wed, 01 Mar 2023 16:38:08 +0000

by: qqq123456

2 comments

Color fixes.

Sun, 29 Jan 2023 13:29:24 +0000

by: qqq123456

Description}: Just realized I mixed up the files so reuploaded with correct colors. In order to avoid repeated posting, a nude version was additionally generated.

catgirl

Fri, 27 Jan 2023 04:34:03 +0000

by: qqq123456

1 comment

Variation

Thu, 26 Jan 2023 13:25:35 +0000

by: qqq123456

Description}: "You're not that tough after all, are you?" I'm sorry if there are grammar or translation problems. My English is terrible....

8 comments

Hands on the wall

Thu, 26 Jan 2023 10:59:46 +0000

by: qqq123456

2 comments

lying on the table

Thu, 26 Jan 2023 10:26:11 +0000

by: qqq123456

Models

Sun, 22 Jan 2023 18:59:57 +0000

by: qqq123456

Description}: As shown in the picture, you can use different models to generate pictures with different artistic styles... It's a bit like the RTX On / RTX Off meme. If you are interested in changing models or art style, please comment below, thank you. Full size image: https://ibb.co/23FwbZk

7 comments

Fox girl

Sun, 01 Jan 2023 15:33:48 +0000

by: qqq123456

Description}: Just another randomly generated image

3 comments

Comics art style.

Sun, 01 Jan 2023 15:08:39 +0000

by: qqq123456

1 comment

Watercolor art style.

Sun, 01 Jan 2023 15:07:40 +0000

by: qqq123456

Description}: I'm trying to generate images with different art styles.

winter

Sun, 25 Dec 2022 14:53:05 +0000

by: qqq123456

Description}: Winter seems to be colder than usual this year, so it's necessary to keep every part of your body warm.

1 comment

Merry Christmas!

Sun, 25 Dec 2022 14:47:08 +0000

by: qqq123456

Description}: It's officially Christmas!! Santa costumes are so cute, don't you think?

Merry Christmas!

Tue, 13 Dec 2022 22:31:36 +0000

by: qqq123456

Description}: it's almost Christmas time! Happy Holidays everyone! ps: I have encountered some work difficulties, so I may reply to private messages later, thank you for waiting.

2 comments

legs up !

Sun, 11 Dec 2022 12:07:23 +0000

by: qqq123456

Description}: I've tried generating some new poses But I think it is still difficult to generate a work that includes two people... I am still adjusting the details.

2 comments

Dressing room

Sun, 11 Dec 2022 11:14:21 +0000

by: qqq123456

Description}: "What took you so long?" Well...the character should be "raimon natsumi" from "Inazuma Eleven" But obviously it still needs improvement, even hypernetwork training has not improved much. Please let me know your thoughts by commenting below, thanks

3 comments

No whispering

Fri, 09 Dec 2022 15:35:11 +0000

by: qqq123456

Description}: "If you're so desperate to tell your friend the answer, why don't you come up here miss?" What's worse than makeing troubles not just for you, but even for your friends? ----------------- This picture can refer to my previous works: https://animeotk.com/gallery/showphoto.php/photo/79568/title/school-road/cat/500 Thanks to @bob3837 for the inspiration. Also, sorry for my bad English, if there is any mistake, please leave a comment below, thanks.

2 comments

Hold-your-position

Fri, 09 Dec 2022 09:57:07 +0000

by: qqq123456

Description}: Just another version.

4 comments

Hold your position

Fri, 09 Dec 2022 09:31:00 +0000

by: qqq123456

Description}: "If you move again, we'll have to start over." Hi, this time I tried to use stable sifutuon alone to generate a full-body image of a character Although the resolution of the result is not high, I still regard it as a success. I'm not sure if two characters are possible on current hardware, but the new graphics card will probably cost me an arm and a leg, so I'll probably try it in the near future. If you have any comments or suggestions, please leave a comment below. Thanks for watching.

2 comments

Canning

Mon, 05 Dec 2022 15:42:11 +0000

by: qqq123456

Description}: This picture tested waifu-diffusion-v1-4 It can be seen that the effect of waifu-diffusion is not as good as anything v3, but it still has its own style. The red marks were drawn using photoshop Any questions or ideas please leave a comment below, thanks

4 comments

Having a pet

Sun, 04 Dec 2022 05:18:04 +0000

by: qqq123456

Description}: "Are you a cat person or a dog person?" No matter which one you prefer, it is very important to take care of it from head to bottom.

5 comments

School road

Sat, 03 Dec 2022 04:06:25 +0000

by: qqq123456

Description}: "It seems that the new class teacher is not easy to mess with." Inspiration provided by @sendynew I initially wanted to generate a full body map of the two, but my graphics card wasn't powerful enough to do that. Images containing two people require more time to screen the finished product I think you need to use at least rtx-20xx graphics card layer to get the job done efficiently If you have any questions or ideas, please leave a comment below, thank you

5 comments

The first time.

Thu, 01 Dec 2022 15:00:11 +0000

by: qqq123456

Description}: There's always a first time! Whether it's forgetting to bring your homework, or getting your first spanking. The image is generated using stable diffusion

7 comments

The simpler the better

Sat, 26 Nov 2022 13:01:09 +0000

by: qqq123456

Description}: I found that I can use the 3d model of drawing software (such as: "CLIP STUDIO PAINT") to assist in generating pictures..... Maybe he can't generate perfect results, but he does make the process a lot easier

Stable Diffusion 2

Fri, 25 Nov 2022 16:52:05 +0000

by: qqq123456

Description}: Another work using Stable Diffusion If you have any suggestions, please comment below, thank you

4 comments

Stable Diffusion

Fri, 25 Nov 2022 16:29:19 +0000

by: qqq123456

Description}: Image generation using Stable Diffusion Compared to Novel AI, it seems to have a better understanding of the details of the hand Although due to computer hardware, I can't use higher resolution generation But for a completely free software, it does have excellent results

A simple instructions

Tue, 22 Nov 2022 18:02:53 +0000

by: qqq123456

Description}: The character in the picture should be sage from valorant https://gelbooru.com/index.php?page=post&s=view&id=5976275&tags=sage_%28valorant%29+ You can see that for NovelAI, drawing fingers is still a problem ----------------------------------- Besides the pictures, I hope I can briefly mention a few points about my upload: ------generally------- In the future, I will tend to regard pictures as "works" rather than "experimental results", which means that I will only upload pictures that I think meet the quality, rather than uploading when technology has improved, and I will reduce the frequency of uploading, in order to Avoid hand-drawn artists' paintings from being washed out -------theme-------- Also, if you want to suggest any characters or themes, please use "Private Message", "Message" or leave a message under the comment area of my picture to interact with me But please don't make "requests" in any other way as that is forbidden There is no fee for making recommendations But I only check the comments every once in a while, and not every topic is suitable for ai, so please don't keep repeating the same suggestions. (It would be nice to have additional relevant information provided) -------other-------- Finally, drawing of very young characters is prohibited as it violates the rules of the forum. I will adjust the image results according to the topic to make sure he meets the age limit, or avoid related topics ------------------------ If I made a mistake somewhere or you have other insights, please leave a comment below, thank you

4 comments

A quick update 3

Sat, 19 Nov 2022 18:55:23 +0000

by: qqq123456

Description}: Another picture This one mimics the redness caused by the Birch If anyone has suggestions please comment below, thanks

5 comments

A-quick-update 2

Sat, 19 Nov 2022 18:21:32 +0000

by: qqq123456

Description}: another picture This one mimics the redness caused by the belt If anyone has suggestions please comment below, thanks

4 comments

A quick update

Sat, 19 Nov 2022 17:44:38 +0000

by: qqq123456

Description}: Improved the post-production method (Photoshop) to make the picture look more natural,thanks for everyone's support

2 comments

Some things that'll never change

Fri, 18 Nov 2022 16:43:33 +0000

by: qqq123456

Description}: "You Are Never Too Old to Be Spanked" A simple exercise, I hope to learn how to draw with gloss and other tools in the future

3 comments

TESE-02

Fri, 18 Nov 2022 14:06:44 +0000

by: qqq123456

Description}: The direction of the description above is " Ellie (the Last of Us Part 1); Rin (Tenchu)" Since I am very familiar with these two characters, I selected the results based on Internet sources Among them, "Ellie" has relatively more information, but the color of the eyes is not sure whether it will be green or blue (I choose green) And "Rin" has very little relevant information, so I used the method of stitching pictures to generate results....The results seem to be unqualified, and the hand structure seems to be a bit wrong You can find the character information at the link below: Ellie： https://gelbooru.com/index.php?page=post&s=view&id=2830260&tags=ellie_%28the_last_of_us%29+ Rin： https://tenchu.fandom.com/wiki/Rin If anyone has other comments, please comment below, thanks

17 comments

TEST-01

Fri, 18 Nov 2022 13:04:51 +0000

by: qqq123456

Description}: The direction of the description above is " long blonde haired/bare bubble butt/crying/ loli" The effect of AI generating red butts seems a little unstable today.....I think I still have some parts to improve If anyone has other comments, please comment below, thanks

7 comments

Red-behind Test Results

Thu, 17 Nov 2022 19:41:05 +0000

by: qqq123456

Description}: The above picture uses the generation method of "1063634615" ("Strength": 55; "Noise": 0.2; Number of attempts: 3~5 times) From the pictures you can see many details that have been improved If you have other comments or topics or content you want to see, please comment below, thank you references: https://animeotk.com/gallery/showphoto.php/photo/79347/title/about-novelai-s-red-behind/cat/532 Image Source: https://animeotk.com/gallery/showphoto.php/photo/79337/title/facial-expression-improvement/cat/532 https://animeotk.com/gallery/showphoto.php/photo/79321/title/novelai-setting-strength/cat/532

3 comments

Facial expression improvement-detail

Wed, 16 Nov 2022 16:38:20 +0000

by: qqq123456

Description}: As far as past experience is concerned, AI is not very good at drawing tears and sad facial expressions, but this is not a definitive conclusion Maybe it's because my English is not very good, but after many attempts, I finally confirmed that adding the extra facial feature "prompt" can improve the generated results The above images are all generated using the same prompt, but the image on the right (green frame) has added an additional facial feature prompt, and you can clearly see the improvement in the generated results. At present, further research on facial prompt is still needed (mainly because of language barriers), if you have any comments or ideas, please leave a comment below, thank you. Original Image： https://imgur.com/efxU8Mc

3 comments

another version

Wed, 16 Nov 2022 15:34:23 +0000

by: qqq123456

Description}: For those interested, here's another version of the previous example (Quested byMordegon)

5 comments

Facial expression improvement

Tue, 15 Nov 2022 17:19:33 +0000

by: qqq123456

Description}: Here is another example

5 comments

Facial expression improvement

Tue, 15 Nov 2022 17:17:26 +0000

by: qqq123456

Description}: A quick update, use "|" to combine different "Prompt" to generate better facial expressions The following is an example of generation, and the details of the generation method will be released later.

4 comments

novelAI setting Prompt

Mon, 14 Nov 2022 14:38:54 +0000

by: qqq123456

Description}: Full size image: https://imgur.com/a/p4k7OMz First of all, I will pick a suitable painting, I mostly only use an image of a person to avoid mistakes in the limbs of the person. Therefore, in the example, I first used Photoshop, erased the extra characters in the picture, and imported the picture into novelAI (1>>2) ----------------------------------- Then, I can enter the Prompt for the image. Instead of narrating in sentences, I usually just use tags for production, and there is no specific ordering between tags, but most of them contain the following items: 1. Characters (if needed) 2. Features 3. Hairstyle 4. Clothing 5. Posture Corresponding to painting 2., that is 1.N/A (blank) 2. huge ass, 3.purple hair, 4.handcuffs, 5.on back, And adjust the "Strength" setting of the picture This image is finer, so I use a lower Strength Once the setup is complete, I start generating images (For details on Strength, please refer to another article) (Entering "seed" will prevent me from spawning continuously, so I mostly leave it blank.) (2>>3) -------------------------------------------------- ---------------------------------- In the end, I used a Strength of 0.55 to generate eight pictures, and picked out two of the better pictures, they were pictures 3-1 and 3-2. 3-1 has correct hand details 3-2 has the details of the chain, can match with the handcuffs Therefore, I merged the two pictures into one, and got the finished product, and then I can do post-production and complete the painting (For an example of post-production, please refer to another article) (3>>4) -------------------------------------------------- -------------------------------------- Hope the above information can help those in need. Of course, my skills aren't really good, and I don't know how to draw at all. So there are many problems that I can't solve, such as: 1. Inability to draw precise mouth poses 2. Inability to draw poses of more than one person 3. Crying does not match most character poses If you want to add additional characters, you can refer to this article: https://www.pixiv.net/artworks/101979863 Articles about "Strength" and "Post-production": https://animeotk.com/gallery/showphoto.php?photo=79321 Most of the above content is generated by a translator, please notify me in the comment area if there is any error, thank you.

3 comments

NovelAI setting trength

Mon, 14 Nov 2022 12:56:48 +0000

by: qqq123456

Description}: Some friends are interested in AI drawing, so I will briefly explain what I know and attach examples The above pictures are all using prompt: Takagi-san, flat ass, looking at viewer, no panties, no pants From the above you can see the effect of the "strength" setting on the image In the original picture, takagi-san is wearing jeans, but starting from 65% intensity, the jeans are gradually erased by AI, at 85%, the jeans disappear completely, and at 95% the composition of the whole picture changes. Different pictures are suitable for different "Strength", for example, 85% and 95% of this picture are acceptable (I think 85% will not change the art style, so it is more suitable) After generating the image, we can post-process it to match the theme (I only understand the most basic modifications) Such as: picture A (strength 85%, additional tag "blush" ), picture B (strength 90%, additional tag "blush" ) Generally speaking, the finer the painting, the lower the intensity will be suitable. P.S. If there are any mistakes, please let me know in the comments below, and you are welcome to discuss with me if you have other comments Full size image：https://imgur.com/a/CrGWCyL

7 comments

AI mimic 3

Sun, 13 Nov 2022 12:49:18 +0000

by: qqq123456

Description}: This is also a repainted work by novelai In fact, AI can draw pictures with high efficiency, but it is difficult to understand the works of more than one person, so there is only one person in the above works. I'm not sure whether AI works are allowed in this forum, please let me know if there are relevant rules If you have any ideas or insights on AI drawing, please comment below, thanks I did not draw the original image, please visit the author's page for information on the original image: https://animeotk.com/gallery/showphoto.php/photo/26015/title/meru-in-the-corner/cat/939 sorry for my bad english, english is not my native language

7 comments

AI mimic 2

Sun, 13 Nov 2022 12:40:53 +0000

by: qqq123456

Description}: This is also a work repainted by novelai I did not draw the original image, please visit the author's page for information on the original image: https://animeotk.com/gallery/showphoto.php/photo/18612/title/just-random-spanking/cat/897 sorry for my bad english, english is not my mother tongue If there are copyright issues or other mistakes, please let me know by leaving a message below, and I will delete it as soon as possible.

5 comments

AI mimic 1

Sun, 13 Nov 2022 12:35:51 +0000

by: qqq123456

Description}: At present, AI still can't independently draw a painting about spanking...or understand the meaning of "spanking". But AI can still help us draw spanking-related works. Above is an old drawing recreated by NOvelai (after post adjustment) I did not draw the original image, please visit the author's page for information on the original image: https://animeotk.com/gallery/showphoto.php/photo/37200/title/nidalee-s-red-ass/cat/1026 Sorry for my bad english If there are copyright issues or other mistakes, please let me know by leaving a message below, and I will delete it as soon as possible.

7 comments