hacksider

Меняем лицо на вебке по одному фото — Deep-Live-Cam Deep-Live-Cam — опенсорсная программа для подмены лица в реальном времени: на вебке, в стриме или на готовом видео. Работает с одной фотографией. 💻 Что умеет: • Сохраняет движения... hacksider/Deep-Live-Cam run.py

I build an open-source local workflow creator so that you will always have a repeatable result I kept noticing something while using AI agents: I was asking them to figure out the same tasks over and over. The first time, that makes sense. The agent needs to understand what I want and work out how to do it. But by the tenth run, I already know the process. I just want it to follow those steps and finish. Instead, it would plan again, make more tool calls, and sometimes take a different approach—even with the same prompt. The results were often useful, but the extra time and token usage started adding up. So I thought, “Why don’t I let AI build the process once, then save that process as a workflow?” That’s how I started building it. I wanted to describe a task using my existing ChatGPT or Claude subscription, have AI turn it into connected steps, then review and reuse it. If something needs judgment, it can still go through AI. But if the next action is simply “check whether another page exists” or “save these rows to a CSV,” ordinary code can handle that. It grew from there. I added conditions, loops, counters, delays, uploads, forms and API calls. You can watch the active step glow as it runs and see the output appear along the way. Then I wanted to check whether it actually made a difference. I ran a scraper benchmark with 10 paired runs. The prompt-driven MCP agent averaged 40.05 seconds, while the saved native workflow averaged 6.12 seconds—about 6.54× faster. Both finished every run successfully and produced the same expected results. The agent averaged 153,304 model tokens per run, including cached input. The native workflow used zero model tokens during execution, because those particular steps ran entirely as code. There’s a catch worth being clear about: this was a small test using fictional local pages and a prebuilt scraper. It didn’t include the cost of creating the workflow, and workflows that contain AI steps still use tokens. But it showed me that, for a predictable task, I didn’t need an agent to rethink the whole process every time. I’ve made the source available so other people can try it too. It’s free for personal and noncommercial use. So the link is hacksider/MCP-Workflow: MCP Workflow Creator

Github ยังไม่สั่งลบอีกเหรอ hacksider/Deep-Live-Cam

给大家分享个神奇的项目。 Deep-Live-Cam 这个项目快十万星。 一张照片，摄像头一开，实时换脸。 为什么神奇呢，因为他比较古早，方法也很简单，就是两个开源模型串起来，加个摄像头，当然也支持离线视频。 部署没什么门槛，8G显存就能跑，效果还可以，感兴趣可以试下。 hacksider/Deep-Live-Cam t.co

𝐀𝐧 𝐀𝐈 𝐣𝐮𝐬𝐭 𝐩𝐚𝐬𝐬𝐞𝐝 𝐭𝐡𝐞 𝐯𝐢𝐝𝐞𝐨 𝐓𝐮𝐫𝐢𝐧𝐠 𝐭𝐞𝐬𝐭. 𝐇𝐚𝐥𝐟 𝐭𝐡𝐞 𝐩𝐞𝐨𝐩𝐥𝐞 𝐨𝐧 𝐭𝐡𝐞 𝐜𝐚𝐥𝐥 𝐭𝐡𝐨𝐮𝐠𝐡𝐭 𝐢𝐭 𝐰𝐚𝐬 𝐡𝐮𝐦𝐚𝐧. On 1 October a company called Tavus unveiled Griffin.  Not a chatbot. Not a talking avatar.  A full video-call partner that watches your face, hears your tone, interrupts you, laughs at the right moment, and answers in real time. In a blind study, 48% of people talking to it believed there was a person on the other end.  Previous systems managed under 3%. Tavus says it will not release it publicly until it is "safe". Nobody has yet explained what that would mean. 𝐀𝐧𝐝 𝐭𝐡𝐞 𝐨𝐩𝐞𝐧-𝐬𝐨𝐮𝐫𝐜𝐞 𝐰𝐨𝐫𝐥𝐝 𝐢𝐬 𝐧𝐨𝐭 𝐟𝐚𝐫 𝐛𝐞𝐡𝐢𝐧𝐝. Last week a 4-billion-parameter model called OneStreamer, built on Alibaba's Qwen, was released with open weights.  It watches a live video feed, remembers what it saw, and decides on its own when to interrupt you.  It does not make a face.  But Deep-Live-Cam already does the face, free, on a gaming PC.  𝐍𝐨𝐛𝐨𝐝𝐲 𝐡𝐚𝐬 𝐭𝐨 𝐰𝐚𝐢𝐭 𝐟𝐨𝐫 𝐓𝐚𝐯𝐮𝐬 𝐭𝐨 𝐝𝐞𝐜𝐢𝐝𝐞 𝐰𝐡𝐚𝐭 "𝐬𝐚𝐟𝐞" 𝐦𝐞𝐚𝐧𝐬. 𝐓𝐡𝐞 𝐠𝐨𝐨𝐝: A tutor who looks at you while explaining, in any language, at 2am. A support line where you talk to a face, not a menu. Rehearse a job interview or a hard conversation against something that reacts like a person. A late relative's face and voice, if the family wants that. 𝐓𝐡𝐞 𝐛𝐚𝐝: Your "CFO" on a video call approving a transfer. That one already happened with cruder tools and cost a firm $25m. Your "grandson" in tears asking for bail money. A "recruiter" interviewing you to harvest your ID documents. Anyone, as anyone, on any platform, in real time. 𝐖𝐡𝐚𝐭 𝐭𝐨 𝐝𝐨 𝐧𝐨𝐰: Agree a family code word for money or emergencies. Never trust a face on a screen as proof of identity. Treat a video call like a text message: it proves nothing about who sent it. The face on your screen is no longer evidence. Adjust accordingly. Griffin: tavus.io/griffin OneStreamer: MCG-NJU/OneStreamer Deep-Live-Cam: hacksider/Deep-Live-Cam #AI #Deepfakes #CyberSecurity #FraudPrevention #OpenSource hacksider/Deep-Live-Cam

9.7 万星。Deep-Live-Cam 实时换脸，一张照片就够了，摄像头前直接换。 门槛降到几乎为零，滥用的成本也跟着降。视频里那个人是不是本人，得留个心眼了。 hacksider/Deep-Live-Cam