RYOUGI Ilakya (@RY0UGI) header
RYOUGI Ilakya (@RY0UGI) profile picture

RYOUGI Ilakya

@RY0UGI

Disagreement enjoyer. Unpopular opinion. Usually take an opposing stance. I believe I'm right so you don't have to.

Joined October 2009
  • Tweets16.4K
  • Following340
  • Followers153
2.7KMedia
Photo by RYOUGI Ilakya (@RY0UGI)Photo by RYOUGI Ilakya (@RY0UGI)Photo by RYOUGI Ilakya (@RY0UGI)Photo by RYOUGI Ilakya (@RY0UGI)Photo by RYOUGI Ilakya (@RY0UGI)Photo by RYOUGI Ilakya (@RY0UGI)Photo by RYOUGI Ilakya (@RY0UGI)Photo by RYOUGI Ilakya (@RY0UGI)Photo by RYOUGI Ilakya (@RY0UGI)Photo by RYOUGI Ilakya (@RY0UGI)
Z-Image Turbo as the photo generator Flux 2 Klein as the style transfer Totally random prompt. Control by concept slider loras on ZIT and the self train style transfer loras blend on Klein to get this specific style consistently. 80-120 sec for the real-stylize pair on RTX3060.
Image from RYOUGI Ilakya (@RY0UGI)
Image from RYOUGI Ilakya (@RY0UGI)
Image from RYOUGI Ilakya (@RY0UGI)
Image from RYOUGI Ilakya (@RY0UGI)
2
1
5,334
RYOUGI Ilakya retweeted
xuan 👼🏻🐉 (@zekeyxiv) profile picture
xuan 👼🏻🐉@zekeyxiv17h
#FF14ファンフェス
1
1,116
12
3,706
98,641
RYOUGI Ilakya retweeted
Stefan 3D AI (@Stefan_3D_AI) profile picture
Stefan 3D AI
@Stefan_3D_AI12h
Finally! Rigged hair cards from a single image. HairDeck (Tsinghua + CMU + CUHK) takes one character image and builds a stylized haircut as separate 3D hair cards, with bone chains already attached. The hair sways when the head turns. Exports to GLB, or straight onto a VRM avatar. Code and weights are open under Apache 2.0. You get geometry and UVs, no textures yet. Local run wants Linux and CUDA. github.com/hyz317/HairDeck
29
309
44
2,883
109,963
สติสตัง ไปหมดแล้ว
Image from RYOUGI Ilakya (@RY0UGI)
4
RYOUGI Ilakya retweeted
satounbd (@satounbd) profile picture
satounbd
@satounbdOct 27
イラストから3Dイラストアニメーション作成の進捗 イラストからLoRAで360度動画を生成→ 基準角度ごとにSee-throughでイラスト分解→ さらに髪の毛を房ごとに画像生成で分解→ パーツをメッシュ化してIDを付与。3D空間で一意にする→ パーツに分解したイラストを対応づけ→ 動かす 本当、See-throughニキが神すぎる
27
348
25
3,403
173,277
RYOUGI Ilakya retweeted
Philtomato 🍅 (@otagoth) profile picture
Philtomato 🍅
@otagothOct 26
When you're an artist but you own a cat.
66
1,010
31
17,659
288,024
RYOUGI Ilakya retweeted
DAILY (@dailycultures) profile picture
DAILY
@dailyculturesOct 26
Uma estudante de 16 anos criou um aparelho que apaga o fogo usando SOM, sem água ou produtos químicos. A invenção conquistou uma medalha de ouro na Indonésia.
Image from DAILY (@dailycultures)
1,083
13,755
1,624
155,765
13,052,417
RYOUGI Ilakya retweeted
Golf (@igolf) profile picture
Golf@igolfOct 25
ไอพวกฝูงลิงท่อง go woke go broke เพิ่งหน้าแหกทั้งโลกตอน The Odyssey ไปเอง (ต้องมาอ้อมแอ้มตบไหล่ปลอบกันทีหลังว่ะก็ไม่โว้คนะ) นี่กลับมาท่องคำเดิมอีกแล้ว พวกมึงนี่ก็ความจำสั้นกันดีเนาะ
36
57
1,036
Endwalker theme + Matoya's Cave theme = Y'stola theme
DISSIDIA DUELLUM FINAL FANTASY (EN) (@DDFF_EN) profile pictureDISSIDIA DUELLUM FINAL FANTASY (EN)
@DDFF_ENOct 24
[#DDFF Livestream Latest Info #3] "Tea on the way home? One would think we were students." ✦ ──── ☪︎ ⊰ ⋆ ⊱ ☪︎ ──── ✦ 　　　　　 2026.10.27 　　　　Y'shtola Arrives! ✦ ──── ☪︎ ⊰ ⋆ ⊱ ☪︎ ──── ✦ youtu.be/EwrQy1PcPz0 #DISSIDIA #FF14
42
RYOUGI Ilakya retweeted
maya 🎹🤎 (@MayaCoffeeElf) profile picture
maya 🎹🤎@MayaCoffeeElfOct 22
Final Fantasy XIV - Answers 🗝️
15
112
4
733
11,391
Torgal Cannon #FF16 #FF8
Image from RYOUGI Ilakya (@RY0UGI)
49
It's not an "if". This is the team behind the The 4 Heroes of Light released at the same time as FF13. Then it became Breavely Default, Octopath Traveller then this. The name wasn't FF but it's branches from FF to begin with. And evolved through the years. #finalfantasyresonance
Image from RYOUGI Ilakya (@RY0UGI)
48
RYOUGI Ilakya retweeted
青龍聖者 (@bdsqlsz) profile picture
青龍聖者
@bdsqlszOct 21
Some people have already attempted to train DLSSNR and have successfully fine-tuned it for use in games.
Image from 青龍聖者 (@bdsqlsz)
Image from 青龍聖者 (@bdsqlsz)
Image from 青龍聖者 (@bdsqlsz)
青龍聖者 (@bdsqlsz) profile picture青龍聖者
@bdsqlszOct 20
Release complete DLSSNR unpack, fine-tuning, and pack code. All you need is 16GB of VRAM to fine-tune your own DLSSNR model.(Lora just 8GB in 1024X1024)
Image from 青龍聖者 (@bdsqlsz)
7
14
6
252
28,921
What in the KH is this!?
𝕄𝕖𝕣𝕣𝕚𝕨𝕖𝕒𝕥𝕙𝕖𝕣 🌙 (@HaMerriweather) profile picture𝕄𝕖𝕣𝕣𝕚𝕨𝕖𝕒𝕥𝕙𝕖𝕣 🌙@HaMerriweatherOct 20
So Infinity Nikki is actually a 2D world
19
The mods community would be wild.
青龍聖者 (@bdsqlsz) profile picture青龍聖者
@bdsqlszOct 20
Release complete DLSSNR unpack, fine-tuning, and pack code. All you need is 16GB of VRAM to fine-tune your own DLSSNR model.(Lora just 8GB in 1024X1024)
Image from 青龍聖者 (@bdsqlsz)
1
42
RYOUGI Ilakya retweeted
Ostris (@ostrisai) profile picture
Ostris
@ostrisaiOct 19
I made this 6-7 years ago using some style transfer methods I was working on at the time. I spent months of tinkering to get this level of temporal consistency. We have come so far so fast. I cannot even imagine where things will be in another 6 years at this pace.
10
2
109
4,609
Diversity done right. "Beauty come in many shape and forms." Not "Beauty is a sin."
Image from RYOUGI Ilakya (@RY0UGI)
18
RYOUGI Ilakya retweeted
AI Search (@aisearchio) profile picture
AI Search
@aisearchioOct 17
The choice of demo images from the MOSS Image team 🤣
Image from AI Search (@aisearchio)
Image from AI Search (@aisearchio)
32
11
4
578
31,466
RYOUGI Ilakya retweeted
Wildminder (@wildmindai) profile picture
Wildminder
@wildmindaiOct 16
We’ve already been eating good with all the small/ fast models dropping lately, and... a new one. 8.56B MOSS-Image-Base- built for insane control. uses JSON layouts and bounding boxes to place text, objects exactly where you want them. + native ComfyUI support Open source is moving at crazy speeds right now. huggingface.co/OpenMOSS-Team…
Image from Wildminder (@wildmindai)
Image from Wildminder (@wildmindai)
Image from Wildminder (@wildmindai)
ModelScope (@ModelScope2022) profile pictureModelScope
@ModelScope2022Oct 16
Meet MOSS-Image-Base: an 8.56B open-weight text-to-image foundation model built for more controllable generation. 🤖 modelscope.ai/collections/op… 🧩 More than natural-language prompts: Generate from descriptive prompts, tag sequences, or JSON layouts that specify text and object placement. 📝 Layout-aware images: Control captions, scene elements, and bounding boxes for posters, whiteboards, diagrams, and other structured visuals. 🔧 Built for customization: As the base checkpoint in the MOSS-Image family, it is designed for research, fine-tuning, and workflow-specific adaptation. 🖥️ Ready for local workflows: Includes Diffusers support, a Gradio interface, CPU offloading options, and a native ComfyUI plugin. 📜 Model weights and inference code are released under Apache 2.0.
Image from ModelScope (@ModelScope2022)
2
21
2
253
15,620
RYOUGI Ilakya retweeted
Studio Wrong (@studiowrong) profile picture
Studio Wrong
@studiowrongOct 16
Beach Therapy 🏖️ Full Animation on YouTube!
22
1,648
24
12,068
171,214
อ่านภาษาอังกฤษไม่ออก เสือกอยากจะทะเลาะกับต่างชาติ แล้วก็โทษตัวแปรภาษา ปัญญามีแค่นี้
Image from RYOUGI Ilakya (@RY0UGI)
76
Load more