Download
1 variant available
bf16 SafeTensor
SinoX_QW2.1_Anything2Real_v1.0_r64.safetensors
BF16, good balance • 304.07 MB
Verified: 2 days ago

346
342
License:
Qwen Research License AgreementQwen is licensed under the Qwen RESEARCH LICENSE AGREEMENT, Copyright (c) 2026 Hangzhou Tongyi Laboratory Technology Co., Ltd. All Rights Reserved.
模型体验地址/Model Playground(邀请码Invite Code: rh-v1603)https://www.runninghub.ai/ai-detail/2104933723808878594?inviteCode=rh-v1603
通过邀请链接注册领1000RH币,每天登录还可以领取100RH币,可以免费畅快跑图了!
Sign up through the invite link to get 1,000 RH coins! Plus, log in daily to claim 100 more RH coins. You can now generate images for free!
My Youtube:https://www.youtube.com/@mugicha-AI
My Discord:https://discord.gg/zqBTpyxwX
如非作者同意禁止搬运。Please do not redistribute without consent.
把二次元,变成真实
一个面向 QwenImage2.1 的风格转换 LoRA。丢一张动漫图进去,出来的是有真实皮肤质感、真实光影的写实影像。
触发词:Convert the image to realistic photograph
Turn 2D anime into real.
A style-transfer LoRA designed for QwenImage2.1. Drop in an anime image, and out comes a photorealistic image with real skin texture and realistic lighting.
Trigger Word:Convert the image to realistic photograph
它能做什么
真实皮肤质感 毛孔、绒毛、细腻的肤色过渡全部保留。不做磨皮式的塑料脸——这是大多数风格转换模型最容易翻车的地方。
光影重塑 把动漫的平涂光照还原成方向光、柔和阴影、真实高光。立体感是「打光打出来的」,不是画出来的。
发丝与织物 头发从色块变成有层次的发丝,衣料从色块变成有纹理的布料。丝质反光、棉麻质感、湿发的重量感都能出来。
构图稳定 转换过程中保持原图的姿态、取景与画面结构。不乱改姿势、不擅自换背景、不重画五官。你要的是转换,不是重绘。
What it can do:
Realistic skin texture — Pores, fine facial hair, and delicate skin tone transitions are all preserved. It doesn't produce the plastic, airbrushed faces that most style-transfer models tend to create — that's where they usually fail.
Lighting reconstruction — It restores the flat shading of anime into directional light, soft shadows, and realistic highlights. The sense of depth comes from actual lighting, not from painting.
Hair strands and fabrics — Hair transforms from color blocks into layered strands, and clothing goes from flat color patches into textured fabric. Silk-like reflections, cotton-linen textures, and the weight of wet hair all come through.
Stable composition — During the conversion, it maintains the original pose, framing, and overall composition. It doesn't arbitrarily change poses, swap backgrounds, or redraw facial features. You want a conversion, not a repaint.
