Sanster @sfjccz Twitter profile

Pinned Tweet

Sanster

10 months

My first macOS App: #OptiClean is officially released! This is an AI application that can cleanup your picture in seconds, the model is completely run locally. 7 day free trial, permanent use after purchase. - Website：

2

4

56

Last Seen Profiles

@samasmaverna

@gestosierra

@Grunzfuss

@haraldssoninho7

@hao_says_hao

@SunsplashMix

@MDMALEKVAI23729

@ibakarachi

@imluvlee

@mercarihallo

@BinorRaja

@BhopalSinghBdla

@pengen_stw

@galery_basah10

@GCUultrasound

@TheRafiqueAhmad

@shoyocel

@eman619rv

@Vicky_9967

@Lulu_du_13

@dul_turkporno

@Win__ston

@fixwell_jp

@gsc_hama

@mayui026

@hoasenfood

@OpenTableUKI

@MichaelPhelps

@bandar_nar9815

@lilmisshanban

@CoinFieldEX

@DRT_Transport

@yolkzheltok

@galery_basah10

@WIPO

@Jf6lKOb2A6Cfu6n

Sanster

@sfjccz

1 year

@Fenng 开源，免费

12

96

409

Sanster

@sfjccz

1 year

#tldream 终于有时间做个 demo 视频了

Sanster

@sfjccz

1 year

Introducing #tldream , a tiny little diffusion drawing app powered by @tldraw and #ControlNet .

2

13

88

12

37

227

Sanster

@sfjccz

8 months

@foxshuo 挺有创意的

1

213

Sanster

@sfjccz

1 year

@nishuang 想要自己部署的推友可以试下我的开源版本，有更好的绘画体验，并且可以使用 HuggingFace 上的任意 Stable Diffusion 模型

4

32

140

Sanster

@sfjccz

2 years

@fei1989long @lili8lili8lili @lyricwai

GitHub - Sanster/IOPaint: Image inpainting tool powered by SOTA AI Model. Remove any unwanted...

Image inpainting tool powered by SOTA AI Model. Remove any unwanted object, defect, people from your pictures or erase and replace(powered by stable diffusion) any thing on your pictures. - Sanster...

github.com

0

35

100

Sanster

@sfjccz

1 year

Introducing #tldream , a tiny little diffusion drawing app powered by @tldraw and #ControlNet .

2

13

88

Sanster

@sfjccz

11 months

@himself_65 要不是我放过我就信了

1

0

85

Sanster

@sfjccz

1 year

可以在 colab 上使用免费的 GPU 跑：

tldream.ipynb

Colaboratory notebook

colab.research.google.com

Sanster

@sfjccz

1 year

#tldream 终于有时间做个 demo 视频了

12

37

227

2

11

81

Sanster

@sfjccz

1 year

@decohack 后期要修图片的话可以考虑试试我的项目 Lama Cleaner

5

13

71

Sanster

@sfjccz

2 years

@coolXiao 图像智能修复

2

10

73

Sanster

@sfjccz

2 years

@lili8lili8lili @lyricwai 讲得挺好的，特别是现在人工智能的出现，使得修改图片越来越简单了

4

8

62

Sanster

@sfjccz

1 year

Segment Anything & Inpainting anything in #LamaCleaner

AI at Meta

@AIatMeta

1 year

Today we're releasing the Segment Anything Model (SAM) — a step toward the first foundation model for image segmentation. SAM is capable of one-click segmentation of any object from any photo or video + zero-shot transfer to other segmentation tasks ➡️

143

2K

7K

8

12

64

Sanster

@sfjccz

11 months

我的第一款 macOS app 准备 TestFlight 测试啦，这是一款可以帮你轻松清理、修正图片的 AI 应用，模型完全本地执行（Apple Silicon only）。TestFlight 地址

7

10

56

Sanster

@sfjccz

1 year

4月份发布了 #LamaCleaner 1.0 版本，还集成了 Segment Anything，油管有主播推荐，涨了一波 Revenue。我自己的期望一直是放得很低，毕竟是开源的项目，现在投入的时间也很少。下一个版本计划加上 ControlNet 1.1 的 inpainting 模型，以及 issue 里的需求汇总。BTW Star 破万啦🎉

1

3

52

Sanster

@sfjccz

4 years

@teemo1209 万科是吧。。。这么巧的么，我是左边的😂

7

46

Sanster

@sfjccz

1 year

10天的收入，比我一开始设想的多，现在在考虑把 windows 安装包设成固定收费，取巧地说还算是“免费的”因为会还是可以用 pip 免费安装

3

1

38

Sanster

@sfjccz

1 year

@liuyi0922 实现了T+1沟通，限流以后的回复第二天才能看到

0

32

Sanster

@sfjccz

2 years

@tinyfool 好像黑客帝国的人体电池😂

1

0

29

Sanster

@sfjccz

2 years

@waylybaye 日本的独立开发者基于 vscode(或者是 atom, 记不清了) 改的笔记软件，订阅制收费，好像收入也不错

5

2

28

Sanster

@sfjccz

10 months

看效果感觉很强，看了下公司介绍，做图片转 SVG 已经 15 年了，这两年加上深度学习模型效果起飞，没有前面10几年的积累很难做这么好吧

千寻 🌸

@Crypto_QianXun

10 months

这个图片处理的工具，只要你用过就会立马路转粉😊 能快速轻松地将 JPEG 和 PNG 位图转换为 SVG 矢量，可无限量放大，目前支持的最大图像像素大小为2万像素，不用登录，还能免费使用 🔗

17

247

736

0

6

28

Sanster

@sfjccz

1 year

With absolutely no experience in Swift, I was able to accomplish this with the help of ChatGPT! This is a completely native MacOS app, from the UI to the model, with no JS/Python/PyTorch involved, running incredibly fast on the GPU of the M2 chip.

5

25

Sanster

@sfjccz

1 year

终于发完了！！这个版本改动太多了，每次想发总能发现个小问题，最后准备发版的物料也花了好多时间

Sanster

@sfjccz

1 year

#LamaCleaner 1.0 release. More plugins for your image post-processing cleaning: - RemoveBG: Remove images background - RealESRGAN: Super Resolution - GFPGAN: Face Restoration - RestoreFormer: Face Restoration Other features 👇🧵

2

6

17

0

2

28

Sanster

@sfjccz

2 years

@waylybaye 这个项目貌似挺活跃的

1

26

Sanster

@sfjccz

9 months

完全不知道为什么某一天会卖的多，然后就会觉得世界好大，互联网好神奇

9

2

24

Sanster

@sfjccz

9 months

这个号挺好的，打算从里面找素材定期发一发，app 还是要有推广的啊，这两天都没什么流量 #OptiClean

Photoshop Tricks!

@gisellaesthetic

1 year

New trick

83

5K

33K

4

2

22

Sanster

@sfjccz

1 year

@hzlzh 生成完了图片需要后处理？可以试试我的项目

3

4

21

Sanster

@sfjccz

2 years

@Larry_LiDev @PenngXiao 可以完全本地部署的版本，支持 gpu 和 cpu

GitHub - Sanster/IOPaint: Image inpainting tool powered by SOTA AI Model. Remove any unwanted...

Image inpainting tool powered by SOTA AI Model. Remove any unwanted object, defect, people from your pictures or erase and replace(powered by stable diffusion) any thing on your pictures. - Sanster...

github.com

1

3

22

Sanster

@sfjccz

10 months

标了两天数据🥲

Sanster

@sfjccz

11 months

我的第一款 macOS app 准备 TestFlight 测试啦，这是一款可以帮你轻松清理、修正图片的 AI 应用，模型完全本地执行（Apple Silicon only）。TestFlight 地址

7

10

56

2

1

21

Sanster

@sfjccz

2 years

@willmcgugan Cleanup your image with powerful AI model.

1

3

19

Sanster

@sfjccz

1 year

@HiTw93 感谢推荐🙏 HuggingFace 的 demo 是 LaMa 模型，除此之外还支持使用 Stable Diffusion 来做擦除，结果更逼真

2

4

19

Sanster

@sfjccz

1 year

#LamaCleaner 1.0 release. More plugins for your image post-processing cleaning: - RemoveBG: Remove images background - RealESRGAN: Super Resolution - GFPGAN: Face Restoration - RestoreFormer: Face Restoration Other features 👇🧵

2

6

17

Sanster

@sfjccz

1 year

Midjourney 做了个图标✨ #LamaCleaner

Sanster

@sfjccz

1 year

4月份发布了 #LamaCleaner 1.0 版本，还集成了 Segment Anything，油管有主播推荐，涨了一波 Revenue。我自己的期望一直是放得很低，毕竟是开源的项目，现在投入的时间也很少。下一个版本计划加上 ControlNet 1.1 的 inpainting 模型，以及 issue 里的需求汇总。BTW Star 破万啦🎉

1

3

52

2

5

17

Sanster

@sfjccz

8 months

我本来还以为是调api，原来都是WebAssembly，还和 chrome 团队深度合作优化，好强

Josh W. Comeau

@JoshWComeau

8 months

😮 So friggin’ cool that *Photoshop* can now run entirely in-browser: @addyosmani has a great write-up about how this is possible:

19

68

461

1

6

18

Sanster

@sfjccz

11 months

咨询了一个技术问题，获得了解答，还收获了关注！开心🥳

0

17

Sanster

@sfjccz

9 months

@waylybaye 划重点：百万付费👍

0

16

Sanster

@sfjccz

2 years

@waylybaye 推荐一个弱监督的标注工具，特点是可以写代码来辅助标注（我没实际用过，看推上的评价还挺好的。。）

0

3

14

Sanster

@sfjccz

8 months

#OptiClean 是一款专为 macOS 开发的图片清理应用，可以快速、方便地清理图片上的瑕疵，移除不需要的物体或者人物。目前开启双 11 & 黑五打折活动， 50% 优惠。

OptiClean: Cleanup your picture in seconds

OptiClean is an AI-powered photo retouch app that can remove unwanted people, objects, text, logos, or watermarks from your images.

opticlean.io

0

2

15

Sanster

@sfjccz

5 years

居然还可以这么玩

Peco

@Switch_movie_SS

5 years

マスターブーメラン #ゼルダの伝説 #BreathoftheWild #NintendoSwitch

2

42

176

0

4

9

Sanster

@sfjccz

1 year

又把摊子铺大了点，准备自己训一个模型，数据量翻倍（千万级），再加上 AI生成的图，祝我好运吧

Sanster

@sfjccz

1 year

With absolutely no experience in Swift, I was able to accomplish this with the help of ChatGPT! This is a completely native MacOS app, from the UI to the model, with no JS/Python/PyTorch involved, running incredibly fast on the GPU of the M2 chip.

5

25

2

0

12

Sanster

@sfjccz

2 years

@syhily 可以试试我的项目，能去水印

GitHub - Sanster/IOPaint: Image inpainting tool powered by SOTA AI Model. Remove any unwanted...

Image inpainting tool powered by SOTA AI Model. Remove any unwanted object, defect, people from your pictures or erase and replace(powered by stable diffusion) any thing on your pictures. - Sanster...

github.com

1

0

12

Sanster

@sfjccz

1 year

三月份的用户“捐赠”收入，下一个项目有两个候选项，GPT 的客户端或者在线的 AI 图片处理服务，纠结下搞哪个…每天只有睡前一小时能干活🥲

0

11

Sanster

@sfjccz

1 year

玩了一下 demo

Meta Open Source

@MetaOpenSource

1 year

We’re excited to announce the release of Animated Drawings, a first-of-its-kind #OpenSource project of annotated amateur drawings aimed at helping researchers easily create their own drawing-to-animation experiences or products. Learn more about the tool:

18

156

599

0

1

11

Sanster

@sfjccz

2 years

@disksing 但是最后能跑🤤

1

12

Sanster

@sfjccz

3 years

@tinyfool 分享一个也是可以本地部署「Inpaining」工具，地址

1

4

12

Sanster

@sfjccz

1 year

#tldream A flamingo

1

2

12

Sanster

@sfjccz

7 months

@mcpba @OfflineHelper 让大家充分意识到了身体健康的重要性，大师

1

10

Sanster

@sfjccz

2 years

@yihong0618 感觉他自己还是会看的，只是为了保号吧，毕竟粉丝也不少

1

0

8

Sanster

@sfjccz

1 year

#tldream A penguin

0

1

10

Sanster

@sfjccz

1 year

@tinyfool 造型和我的电脑支架有点像

1

9

Sanster

@sfjccz

6 years

安装包 19MB，不是 JS 写的，保持了 Sublime Text 的简洁风格

Linuxtoy 🐧

@linuxtoy

6 years

Sublime Merge：来自 Sublime Text 编辑器作者的新作，支持 Linux、macOS、以及 Windows 的跨平台 Git 客户端，包括三路 (three-way) 合并工具、并排 diff、语法加亮等不错的功能。()

4

78

207

1

5

10

Sanster

@sfjccz

3 years

@lomyrjo

1

0

8

Sanster

@sfjccz

2 years

#lamacleaner - New UI - Zoom/Pan

1

9

Sanster

@sfjccz

1 year

@waylybaye macos版可以不走 app store 吧

1

0

9

Sanster

@sfjccz

1 year

我来总结一下：训练最小模型的显存需求降了一倍，从 64 块 80G A100，降低到了 32 块

GitHub Projects Community

@GithubProjects

1 year

0

4

9

1

5

9

Sanster

@sfjccz

1 year

又上 GitHub Trending 啦，虽然 Trending 的逻辑有点奇怪，里面居然没有 AutoGPT🤔

0

9

Sanster

@sfjccz

2 years

Grammarly 超好用啊！！鬼知道我以前发了多少语法错误的英文。。

0

3

6

Sanster

@sfjccz

2 years

@nishuang 感谢推荐。如果安装不喜欢安装的话，有一个在线的服务可以使用（不是我开发的）。不过免费版本只能处理 720p 的图片，开发 lama-cleaner 的初衷就是想支持高清图片，以及能够简单地在自己的服务器上部署。

1

9

Sanster

@sfjccz

8 months

开心，有种在旧裤子里摸到钱的感觉

1

0

7

Sanster

@sfjccz

9 months

更新了一版官网，测试下twitter card

OptiClean: Cleanup your picture in seconds

OptiClean is an AI-powered photo retouch app that can remove unwanted people, objects, text, logos, or watermarks from your images.

opticlean.io

3

0

8

Sanster

@sfjccz

10 months

@lencx_ 有人专门注册GPT4的账号的，300块可以买120刀的额度，每个月月底过期，我也不知道咋搞的

4

0

7

Sanster

@sfjccz

2 years

@ProductHunt Open Source Inpainting Tool

GitHub - Sanster/IOPaint: Image inpainting tool powered by SOTA AI Model. Remove any unwanted...

Image inpainting tool powered by SOTA AI Model. Remove any unwanted object, defect, people from your pictures or erase and replace(powered by stable diffusion) any thing on your pictures. - Sanster...

github.com

Sanster

@sfjccz

2 years

这次在 reddit 上发了视频的演示，反响还不错，但还是没人请喝咖啡🥲

1

2

0

1

8

Sanster

@sfjccz

1 year

每次想放放之前的项目，总有新东西出来…这模型比我之前用的交互式分割模型效果好多了，今天就加上！ #LamaCleaner

AI at Meta

@AIatMeta

1 year

Today we're releasing the Segment Anything Model (SAM) — a step toward the first foundation model for image segmentation. SAM is capable of one-click segmentation of any object from any photo or video + zero-shot transfer to other segmentation tasks ➡️

143

2K

7K

0

2

6

Sanster

@sfjccz

9 months

@silver_valyrian Removed using my macOS app:

1

3

8

Sanster

@sfjccz

1 year

@khbzhz 截图是 @lmsqueezy ，卖的软件是我的开源项目

GitHub - Sanster/IOPaint: Image inpainting tool powered by SOTA AI Model. Remove any unwanted...

Image inpainting tool powered by SOTA AI Model. Remove any unwanted object, defect, people from your pictures or erase and replace(powered by stable diffusion) any thing on your pictures. - Sanster...

github.com

1

3

8

Sanster

@sfjccz

11 months

这个 Figma 插件好强啊，操作都在客户端执行，看生成的过程感觉是个小的 diffusion 模型？

Yi Shen

@pissang1

11 months

🥳🥳My Figma plugin "Vector to 3D" has been published! It can convert any vectors to 3D with one simple click. And with its high-quality ray-tracing renderer, you can create beautiful 3D icons, text, or other 3D-style images without being a 3D specialist

32

379

2K

1

7

Sanster

@sfjccz

1 year

@Jimmy_JingLv 用的这个 @screenstudioapp

1

0

6

Sanster

@sfjccz

10 months

@xiqingongzi 看到「不是浪费我们的热情吗？」有点血压升高

0

7

Sanster

@sfjccz

1 year

太牛了，卖 notion 模版一个月 10 万美元…我自己是从来没买过 notion 的模版

Thomas Frank

@TomFrankly

1 year

It's been just over a month since Gumroad increased their prices. It has also been just over a month since I switched to @lmsqueezy for selling my @NotionHQ templates. I'm finally ready to share my findings, along with a massive free tool I've been working on.

46

20

432

2

3

6

Sanster

@sfjccz

3 years

@laixintao 羡慕大内存，我们48核机器才配了128G

0

7

Sanster

@sfjccz

1 year

@design_hacking …这个要怎么证明版权是他们的呢🤔

1

0

7

Sanster

@sfjccz

10 months

#OptiClean 发布一周勉强快回本了，看了下内购率，确实很低（Free Trial是用 0 元内购实现的，说明很多人被劝退了）。目前想的方案是第一次打开提供一张内置图片可以体验，如果想在自己的图片上体验就需要开始 Free Trial

2

0

7

Sanster

@sfjccz

2 years

@liuyi0922 居然只有一个开发，我终于明白什么叫顶梁柱了

0

7

Sanster

@sfjccz

1 year

Image inpainting tool powered by SOTA AI Model.

GitHub - Sanster/IOPaint: Image inpainting tool powered by SOTA AI Model. Remove any unwanted...

Image inpainting tool powered by SOTA AI Model. Remove any unwanted object, defect, people from your pictures or erase and replace(powered by stable diffusion) any thing on your pictures. - Sanster...

github.com

0

2

7

Sanster

@sfjccz

10 months

有一段时间我一度以为这两位推友是一个人😅

0

6

Sanster

@sfjccz

1 year

Stable Diffusion can do "normal" inpainting! And it's good!! Just use the magic prompt "background".

1

0

6

Sanster

@sfjccz

1 year

M2 的 GPU 好强，不，应该说是 coreml 好强，pytorch 即使设置了 device=mps 也没那么快。跑通基础流程还只是第一步，想想还有一大堆事情要做啊🫠

Sanster

@sfjccz

1 year

With absolutely no experience in Swift, I was able to accomplish this with the help of ChatGPT! This is a completely native MacOS app, from the UI to the model, with no JS/Python/PyTorch involved, running incredibly fast on the GPU of the M2 chip.

5

25

2

1

6

Sanster

@sfjccz

1 year

由 replicate 开发的在线服务，可以直接在线体验，前端代码是开源的：

GitHub - replicate/scribble-diffusion: Turn your rough sketch into a refined image using AI

Turn your rough sketch into a refined image using AI - replicate/scribble-diffusion

github.com

0

6

Sanster

@sfjccz

2 years

@yetone 完全被聊天内容吸引了，以至于想在对话中找出利用前端 bug 提取公积金的方法。。

0

6

Sanster

@sfjccz

11 months

更多效果演示以及使用文档可以参见官网

OptiClean: Cleanup your picture in seconds

OptiClean is an AI-powered photo retouch app that can remove unwanted people, objects, text, logos, or watermarks from your images.

opticlean.io

1

0

4

Sanster

@sfjccz

3 years

@cc_aex #再也不停立体车库

1

5

Sanster

@sfjccz

10 months

@waylybaye 这两个放一起挺搭的，单独看的话我更喜欢现在的😅

0

6

Sanster

@sfjccz

1 year

@Suhail is good

GitHub - visual-layer/fastdup: fastdup is a powerful free tool designed to rapidly extract valuable...

fastdup is a powerful free tool designed to rapidly extract valuable insights from your image & video datasets. Assisting you to increase your dataset images & labels quality and re...

github.com

1

0

6

Sanster

@sfjccz

1 year

不知道为什么很多人应该都还没用过就在那里喷，截图是官方的 onboard message，写的还是很实在的

外汇交易员

@myfxtrader

1 year

复旦团队发布国内首个类ChatGPT模型MOSS，邀公众参与内测。MOSS可执行对话生成、编程、事实问答等一系列任务，打通了让生成式语言模型理解人类意图并具有对话能力的全部技术路径。 *ChatGPT难道这么简单吗？大家三下五除二都搞出来了😂

90

32

198

1

2

6

Sanster

@sfjccz

11 months

@thecalicastle 第一眼整体感觉我喜欢左边的

1

0

4

Sanster

@sfjccz

9 months

上午提交了新版本审核，然后通过搜索 @PodwiseHQ + @cleanclipapp 找了几个 AI 导航站提交了下 list 申请。感觉这些 AI 导航站就像新时代的门户网站。。

2

0

5

Sanster

@sfjccz

9 months

@real_kai42 所以当初 huggingface 选择了 gradio……

2

0

5

Sanster

@sfjccz

1 year

@kevinzhow 怪不得 star 了我的项目，是准备自己训识别模型么？

2

0

5

Sanster

@sfjccz

1 year

@zeke Too many requests...

1

0

4

Sanster

@sfjccz

1 year

@YwhyWaiY 能直接在手机上跑确实很强，我开源了一个类似的桌面 AI 软件，有兴趣可以试试

2

0

4

Sanster

@sfjccz

1 year

#LamaCleaner I just learned that you can convert an arbitrary SD model into an inpainting model, here's anything 4.0 inpainting result.

0

5

Sanster

@sfjccz

2 years

@passluo 被收购的三个都听过，收购的那个没听过😂

0

4

Sanster

@sfjccz

7 months

感受到了一点点🤏黑五

2

0

5

Sanster

@sfjccz

2 years

Try it yourself using free open-source inpainting tool

GitHub - Sanster/IOPaint: Image inpainting tool powered by SOTA AI Model. Remove any unwanted...

Image inpainting tool powered by SOTA AI Model. Remove any unwanted object, defect, people from your pictures or erase and replace(powered by stable diffusion) any thing on your pictures. - Sanster...

github.com

0

1

4

Sanster

@sfjccz

11 months

@nebula_moe github private repo 免费的…

0

4

Sanster

@sfjccz

2 years

@syhily

0

4

Sanster

@sfjccz

1 year

@dotey 嗯，集成了 Segment Anything 交互式分割的能力，可以更方便、精准地选择想要 inpainting 的区域

1

0

5

Sanster

@sfjccz

2 years

显著性语义分割，类似的项目有，感兴趣的可以看看，免费开源

GitHub - xuebinqin/U-2-Net: The code for our newly accepted paper in Pattern Recognition 2020:...

The code for our newly accepted paper in Pattern Recognition 2020: "U^2-Net: Going Deeper with Nested U-Structure for Salient Object Detection." - xuebinqin/U-2-Net

github.com