我用Stable Diffusion做电商！

如果你要开淘宝网店，那么在淘宝网页上展示的商品就要漂亮精致，紧紧抓住消费者的心♥！我们可以借助AI作图工具，简单地代替请模特特地搞拍摄的过程啦！这里介绍很简单的利用AI绘画局部逐渐美化女装商品展示图的方法。

我是运营网店的女装店主，我没有钱请模特了。。。

我可以用stable diffusion来初步制作自己的展示商品！

比如我这里要卖这个绿色的淑女裙。（左图）

我尝试了直接拿真人穿的衣服抠出来生成，效果很不好。（右图）

借鉴了一些视频和方法，我总结了一些我觉得实用性较高，也比较简单的步骤。 我觉得局部重绘是比较合适和真实的方法。

真人穿衣服拍照。

拿到穿衣服的比较真实质感的照片。【如果是身材方面有点难处 那就借助美图秀秀or ps吧】

​

ok, 比如我（不具有做模特资质的小美女）穿好了我卖的漂亮衣服摆好pose摆几张啦！

选好底模！

一定要是realistic的，真人照片风格的底模。

我这里选的是很经典的majicmixRealistic_v7。

换头，根据不同平台换头！

比如面向海外市场的，就得换白女头吧。面向中老妇女的，换妈妈头。

ok，我这里换白女头，比如我要放在亚马逊上卖。

操作就是图生图下的局部重绘选项卡下涂抹自己替换的部分。

​

prompts & parameters

breathtaking cinematic photo, masterpiece, best quality, (photorealistic:1.33), blonde hair, silver necklace, carrying a white bag, standing, full body, detailed face,big eyes, detailed hands,

wearing the light green dress, fashion photography, studio light, . 35mm photograph, film, bokeh, professional, 4k, highly detailed . award-winning, professional, highly detailed,

Negative prompt: ugly, disfigured, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry

Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 7, Seed: 3748048683, Size: 362x486, Model hash: 7c819b6d13, Model: majicmixRealistic_v7, VAE hash: c6a580b13a, VAE: vae-ft-mse-840000-ema-pruned.safetensors, Denoising strength: 0.75, Clip skip: 2, Mask blur: 2, ControlNet 0: "Module: none, Model: control_v11p_sd15_openpose [cab727d4], Weight: 1, Resize Mode: Crop and Resize, Low Vram: False, Processor Res: 512, Guidance Start: 0, Guidance End: 1, Pixel Perfect: True, Control Mode: My prompt is more important", Version: v1.6.0

换个手链，不要手表了额

换饰品o(*￣▽￣*)ブ

prompts & parameters

breathtaking cinematic photo, masterpiece, best quality, (photorealistic:1.33), blonde hair, silver necklace, carrying a white bag, standing, full body, detailed face,big eyes, detailed hands,bracelet,

wearing the light green dress, fashion photography, studio light, . 35mm photograph, film, bokeh, professional, 4k, highly detailed . award-winning, professional, highly detailed,

Negative prompt: ugly, disfigured, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry

Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 7, Seed: 3010689187, Size: 362x486, Model hash: 7c819b6d13, Model: majicmixRealistic_v7, VAE hash: c6a580b13a, VAE: vae-ft-mse-840000-ema-pruned.safetensors, Denoising strength: 0.75, Clip skip: 2, Mask blur: 2, ControlNet 0: "Module: none, Model: control_v11p_sd15_openpose [cab727d4], Weight: 1, Resize Mode: Crop and Resize, Low Vram: False, Processor Res: 512, Guidance Start: 0, Guidance End: 1, Pixel Perfect: True, Control Mode: My prompt is more important", Version: v1.6.0

换个珍珠项链

prompts & parameters

breathtaking cinematic photo, masterpiece, best quality, (photorealistic:1.33), blonde hair, carrying a light blue bag, standing, full body, detailed face,big eyes, detailed hands,pearl necklace,

wearing the light green dress, fashion photography, studio light, . 35mm photograph, film, bokeh, professional, 4k, highly detailed . award-winning, professional, highly detailed,

Negative prompt: ugly, disfigured, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry

Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 7, Seed: 1117114026, Size: 362x486, Model hash: 7c819b6d13, Model: majicmixRealistic_v7, VAE hash: c6a580b13a, VAE: vae-ft-mse-840000-ema-pruned.safetensors, Denoising strength: 0.75, Clip skip: 2, Mask blur: 2, ControlNet 0: "Module: none, Model: control_v11p_sd15_openpose [cab727d4], Weight: 1, Resize Mode: Crop and Resize, Low Vram: False, Processor Res: 512, Guidance Start: 0, Guidance End: 1, Pixel Perfect: True, Control Mode: My prompt is more important", Version: v1.6.0

换个好看的包包

prompts & parameters

breathtaking cinematic photo, masterpiece, best quality, (photorealistic:1.33), blonde hair, carrying a cute cartoon bag, standing, full body, detailed face,big eyes, detailed hands,pearl necklace,

wearing the light green dress, fashion photography, studio light, . 35mm photograph, film, bokeh, professional, 4k, highly detailed . award-winning, professional, highly detailed,

Negative prompt: ugly, disfigured, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry

Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 7, Seed: 1117114026, Size: 362x486, Model hash: 7c819b6d13, Model: majicmixRealistic_v7, VAE hash: c6a580b13a, VAE: vae-ft-mse-840000-ema-pruned.safetensors, Denoising strength: 0.75, Clip skip: 2, Mask blur: 2, ControlNet 0: "Module: none, Model: control_v11p_sd15_openpose [cab727d4], Weight: 1, Resize Mode: Crop and Resize, Low Vram: False, Processor Res: 512, Guidance Start: 0, Guidance End: 1, Pixel Perfect: True, Control Mode: My prompt is more important", Version: v1.6.0

换个背景！

抠出模特然后继续生成。

请输入标题

breathtaking cinematic photo, masterpiece, best quality, (photorealistic:1.33), blonde hair, standing, full body, detailed face,big eyes, detailed hands,

wearing the light green dress, fashion photography, studio light, . 35mm photograph, film, bokeh, professional, 4k, highly detailed . award-winning, professional, highly detailed,carrying cute cartoon bag,indoors, window shopping, street, 

Negative prompt: ugly, disfigured, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry

Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 7, Seed: 2728494159, Size: 362x486, Model hash: 7c819b6d13, Model: majicmixRealistic_v7, VAE hash: c6a580b13a, VAE: vae-ft-mse-840000-ema-pruned.safetensors, Denoising strength: 0.75, Clip skip: 2, Mask blur: 2, ControlNet 0: "Module: none, Model: control_v11p_sd15_openpose [cab727d4], Weight: 1, Resize Mode: Crop and Resize, Low Vram: False, Processor Res: 512, Guidance Start: 0, Guidance End: 1, Pixel Perfect: True, Control Mode: My prompt is more important", Version: v1.6.0

我这里抠图中间抠得不是很彻底，最好是ps扣好模板让后画背景，总之这对SD来说小菜一碟。这样就可以做自己的淘宝模特啦。面向不同的人群市场可以做不同的模特，不用请人了哦！

同样地 这种一点点修的方法可以用在视频上：

生成视频的方法有两种：一种是很粗暴地替换

也可以一帧一帧拆开然后利用controlNet+局部重绘功能，每一帧做取模重绘。

好好玩哦！我也是慢慢学，面向实用去学！