Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Impressive work for LTX-2. Face Swap LoRA - Identity-locked video face swapping; - trained on 300+ HQ pairs @ 512x512; - optimized for close-up shots and landscape aspect ratios; Alpha release.

Wildminder

10,441 subscribers

22,907 views • 6 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

SDR to HDR from ComfyUI, I've trained a LoRA over Qwen Edit 2011 based on the principle used in A research I've had the previlige to work on with Naomi Ken Korem for video HDR on LTX. Links for the LoRA and workflow (+ my fun grading node) below.

SDR to HDR from ComfyUI, I've trained a LoRA over Qwen Edit 2011 based on the principle used in A research I've had the previlige to work on with Naomi Ken Korem for video HDR on LTX. Links for the LoRA and workflow (+ my fun grading node) below.

Mohamed Oumoumad

58,774 views • 3 months ago

Face swapping just leveled up. Akool Face Swap just got a major upgrade — designed to make AI video creation faster, smarter, and way more powerful. Now you can process up to 50 files in batch, swap faces directly on GIFs, and benefit from smarter face detection that automatically removes duplicate faces across uploads. We’re also introducing Classic Plus, a new face model, along with Extracted Key Frames for more accurate detection in videos. And for Pro users, you can now keep the original video specs, so your final output stays production-ready. Less manual work. Better results. More creative freedom. Try the new Face Swap on Akool and see the difference. 🚀

Face swapping just leveled up. Akool Face Swap just got a major upgrade — designed to make AI video creation faster, smarter, and way more powerful. Now you can process up to 50 files in batch, swap faces directly on GIFs, and benefit from smarter face detection that automatically removes duplicate faces across uploads. We’re also introducing Classic Plus, a new face model, along with Extracted Key Frames for more accurate detection in videos. And for Pro users, you can now keep the original video specs, so your final output stays production-ready. Less manual work. Better results. More creative freedom. Try the new Face Swap on Akool and see the difference. 🚀

Akool Inc

298,768 views • 5 months ago

Face Replacement and De-aging Demo: VFX artist Doug Hogan built a fully automated face swap workflow inside ComfyUI. Florence 2 → SAM2 → WAN Video. One pipeline. Temporally consistent across every frame. Not a pixel swap. The workflow detects the head, segments it, drives motion from the original performance, and uses a reference image + auto prompt to transfer the identity while preserving lighting and timing. Face replacement. De-aging. All inside ComfyUI. Click the link below for the workflow 👇

Face Replacement and De-aging Demo: VFX artist Doug Hogan built a fully automated face swap workflow inside ComfyUI. Florence 2 → SAM2 → WAN Video. One pipeline. Temporally consistent across every frame. Not a pixel swap. The workflow detects the head, segments it, drives motion from the original performance, and uses a reference image + auto prompt to transfer the identity while preserving lighting and timing. Face replacement. De-aging. All inside ComfyUI. Click the link below for the workflow 👇

ComfyUI

14,833 views • 2 months ago

Ovi is out on Hugging Face Twin Backbone Cross-Modal Fusion for Audio-Video Generation Ovi is a veo-3 like, video+audio generation model that simultaneously generates both video and audio content from text or text+image inputs. Video+Audio Generation: Generate synchronized video and audio content simultaneously Flexible Input: Supports text-only or text+image conditioning 5-second Videos: Generates 5-second videos at 24 FPS, area of 720×720, at various aspect ratios (9:16, 16:9, 1:1, etc)

Ovi is out on Hugging Face Twin Backbone Cross-Modal Fusion for Audio-Video Generation Ovi is a veo-3 like, video+audio generation model that simultaneously generates both video and audio content from text or text+image inputs. Video+Audio Generation: Generate synchronized video and audio content simultaneously Flexible Input: Supports text-only or text+image conditioning 5-second Videos: Generates 5-second videos at 24 FPS, area of 720×720, at various aspect ratios (9:16, 16:9, 1:1, etc)

AK

23,082 views • 10 months ago

From reality to Columbina in seconds. 🕊️✨ Seedance 2 is now available to EVERYONE on MartiniArt — no business plan required (US & Japan supported)! We are currently the ONLY platform that lets you create Seedance 2 videos using a real human face with up to 50% discount! Follow + RT + Comment "seedance 2" and I'll send you: 200 FREE credits, the exact workflow for this video and a tutorial on how to use your own face for Seedance 2! #Seedance2 #MartiniArt

From reality to Columbina in seconds. 🕊️✨ Seedance 2 is now available to EVERYONE on MartiniArt — no business plan required (US & Japan supported)! We are currently the ONLY platform that lets you create Seedance 2 videos using a real human face with up to 50% discount! Follow + RT + Comment "seedance 2" and I'll send you: 200 FREE credits, the exact workflow for this video and a tutorial on how to use your own face for Seedance 2! #Seedance2 #MartiniArt

Astorie AI (prev. MartiniArt)

28,907 views • 4 months ago

In Britain, we tax work more than wealth — unfair and bad for our economy. A 2% tax on assets over £10m would start to fix it. If any unpatriotic billionaires want to run off, fine — sell up to someone who’ll pay and face an exit tax on the way out.

In Britain, we tax work more than wealth — unfair and bad for our economy. A 2% tax on assets over £10m would start to fix it. If any unpatriotic billionaires want to run off, fine — sell up to someone who’ll pay and face an exit tax on the way out.

James Schneider

239,968 views • 9 months ago

Today is Arraignment Day for a man that many people have believed to have been a criminal for decades. While everyone is innocent until proven guilty, I just want to remind you that Trump literally campaigned on "lock her up," and then his DOJ investigated her to no avail for 2 years. Trump supporters chanted "lock her up" for the past 8 years. They literally wanted Trump's political opponents locked up without any due process. So now that their guy has been indicted and will potentially face being locked up through due process, they call it "political bias." I call it hypocrisy.

Today is Arraignment Day for a man that many people have believed to have been a criminal for decades. While everyone is innocent until proven guilty, I just want to remind you that Trump literally campaigned on "lock her up," and then his DOJ investigated her to no avail for 2 years. Trump supporters chanted "lock her up" for the past 8 years. They literally wanted Trump's political opponents locked up without any due process. So now that their guy has been indicted and will potentially face being locked up through due process, they call it "political bias." I call it hypocrisy.

Ed Krassenstein

658,673 views • 3 years ago

Project Launch 🚀: mukh v0.1.14 - The Face Analysis Library Hello Everyone, I am happy to launch my first python package "mukh". Mukh (मुख, meaning "face" in Sanskrit) is a comprehensive face analysis library that provides unified APIs for various face-related tasks. It simplifies the process of working with multiple face analysis models through a consistent interface. --- 📌 Features: 🥸 DeepFake Detector: First Python package featuring an Ensemble of multiple models 🎯 Unified API: Single, consistent API for multiple face analysis tasks like face detection and reenactment 🔄 Model Flexibility: Support for multiple models per task 🛠️ Custom Pipelines: Optimized preprocessing and model combinations --- 📌 Currently Supported Tasks: 1️⃣ Face Detection 2️⃣ Face Reenactment with Source Image and Driving Video 3️⃣ Deepfake Detection for Image and Video 4️⃣ Deepfake Detection Pipeline - Ensemble of multiple models --- 📌 Open Source Contributions: I am opening up multiple tasks in the issues section of mukh including a bunch of `good-first-issues`. My goal is to help a number of beginners get started with their contributions, I will be personally guiding across completing these tasks so keep a check on the issues section. The link to the GitHub repository and detailed documentation is available below 👇 in the reply section, do give it a ⭐️ to support the project! --- Special Mentions in the video: Ayush Chaurasia adi unni hsr hacker house Nimisha Chanda @ElonMastikhor Ctrl+Vibe

Project Launch 🚀: mukh v0.1.14 - The Face Analysis Library Hello Everyone, I am happy to launch my first python package "mukh". Mukh (मुख, meaning "face" in Sanskrit) is a comprehensive face analysis library that provides unified APIs for various face-related tasks. It simplifies the process of working with multiple face analysis models through a consistent interface. --- 📌 Features: 🥸 DeepFake Detector: First Python package featuring an Ensemble of multiple models 🎯 Unified API: Single, consistent API for multiple face analysis tasks like face detection and reenactment 🔄 Model Flexibility: Support for multiple models per task 🛠️ Custom Pipelines: Optimized preprocessing and model combinations --- 📌 Currently Supported Tasks: 1️⃣ Face Detection 2️⃣ Face Reenactment with Source Image and Driving Video 3️⃣ Deepfake Detection for Image and Video 4️⃣ Deepfake Detection Pipeline - Ensemble of multiple models --- 📌 Open Source Contributions: I am opening up multiple tasks in the issues section of mukh including a bunch of `good-first-issues`. My goal is to help a number of beginners get started with their contributions, I will be personally guiding across completing these tasks so keep a check on the issues section. The link to the GitHub repository and detailed documentation is available below 👇 in the reply section, do give it a ⭐️ to support the project! --- Special Mentions in the video: Ayush Chaurasia adi unni hsr hacker house Nimisha Chanda @ElonMastikhor Ctrl+Vibe

Ishan Dutta

13,954 views • 1 year ago

Brian De Palma is usually known for big, operatic set pieces, but in Scarface he opens on a simple close-up of Al Pacino’s face. He later explained why... "I think the most stunning thing about Al is his face. He's the kind of guy that can hold the screen with his face. When you start a movie, you sort of want to give the lead character a very impressive entrance. And that face, that character, the crazy shirt he's wearing, the scar, the way he moves, the way he talks; you just want to really hit them because you're hitting them with something they've never seen before. A lot of this had been reported on the news, but nobody had ever seen it in a movie before. You see these Cuban gangsters and the way they talked, the way they moved. And he just sort of embodied that in that close-up."

Brian De Palma is usually known for big, operatic set pieces, but in Scarface he opens on a simple close-up of Al Pacino’s face. He later explained why... "I think the most stunning thing about Al is his face. He's the kind of guy that can hold the screen with his face. When you start a movie, you sort of want to give the lead character a very impressive entrance. And that face, that character, the crazy shirt he's wearing, the scar, the way he moves, the way he talks; you just want to really hit them because you're hitting them with something they've never seen before. A lot of this had been reported on the news, but nobody had ever seen it in a movie before. You see these Cuban gangsters and the way they talked, the way they moved. And he just sort of embodied that in that close-up."

Gangster Cinema Central

120,340 views • 3 months ago

Seedance 2.5’s most underrated industry upgrade is its 3D white model input rendering — worth a close look for anyone in sci-fi film & VFX. It fixes real pre-production pain points, with clear efficiency wins: - Faster previs iteration: Get atmospheric previews straight from locked layout, no waiting on lighting & texture teams. Cheaper creative revisions. - Lower costs for pitches and concept validation. No wasting senior team hours on bid demos. - Earlier reference for VFX & art teams, extending cross-department parallel work. - Fast turnarounds for secondary shots (distant setups, alternates), saving render time and manpower. Official demo: ~100k-face spacecraft animation + style references → fully rendered footage. Structure and proportions hold steady in dynamic shots, and your original layout stays intact. One clear limit: it is still a generative previs tool, and cannot yet replace industrial physically accurate rendering and standard asset pipelines. It won’t take core lighting & rendering jobs, but it will reshape how previs gets done.

Seedance 2.5’s most underrated industry upgrade is its 3D white model input rendering — worth a close look for anyone in sci-fi film & VFX. It fixes real pre-production pain points, with clear efficiency wins: - Faster previs iteration: Get atmospheric previews straight from locked layout, no waiting on lighting & texture teams. Cheaper creative revisions. - Lower costs for pitches and concept validation. No wasting senior team hours on bid demos. - Earlier reference for VFX & art teams, extending cross-department parallel work. - Fast turnarounds for secondary shots (distant setups, alternates), saving render time and manpower. Official demo: ~100k-face spacecraft animation + style references → fully rendered footage. Structure and proportions hold steady in dynamic shots, and your original layout stays intact. One clear limit: it is still a generative previs tool, and cannot yet replace industrial physically accurate rendering and standard asset pipelines. It won’t take core lighting & rendering jobs, but it will reshape how previs gets done.

Alisa Qian

17,638 views • 1 month ago

SNEAKO absolutely COOKS Tommy Robinson for being cracked out in his latest video and starts impersonating him😂🔥 "Cocaine is a hell of a drug. Advice to Steven Yaxley Lennon, Tommy Rubenstein, everybody knows you're on drugs, put the camera a little furhter away from your face idiot. Why would you show a close up of how f*cked up you are?"

SNEAKO absolutely COOKS Tommy Robinson for being cracked out in his latest video and starts impersonating him😂🔥 "Cocaine is a hell of a drug. Advice to Steven Yaxley Lennon, Tommy Rubenstein, everybody knows you're on drugs, put the camera a little furhter away from your face idiot. Why would you show a close up of how f*cked up you are?"

SNEAKO HQ

148,778 views • 4 months ago

VIDEO | Delhi: Rebel TMC MP Rachna Banerjee, says, "There can never be any rebellion against her (Mamata Banerjee). We have shared a very old relationship with Didi, and that relationship will always remain the same. The respect we have for her will always remain. But one thing is true... everyone says that we received votes because of Didi. That is correct. We got votes because of Didi, because Didi herself is the identity and symbol of the Trinamool Congress. In Trinamool Congress, the face people recognise and trust is Didi’s face, and we all respect her deeply. People voted after seeing Didi, and they voted for Rachna Banerjee not because they were impressed by my face, but because they believed I would be able to work for them. Didi was the face behind it, and after that they gave us their votes with the expectation that I would do good work for the people and for my constituency. There was absolutely no resentment. But I felt that if the Central government and the State government were ruled by the same side, then getting work done would become much easier... something we have not really seen over the past 15 years." "People want us to do something for them, to carry out good work for them. And to make that happen, support from the Centre is very important. So at times, with due respect to Didi, I felt that while staying with her, there were many instances where obstacles were created in our work, and we were unable to do the kind of work we actually wanted to do.” #TMC #RachnaBanerjee (Full video available on PTI Videos -

VIDEO | Delhi: Rebel TMC MP Rachna Banerjee, says, "There can never be any rebellion against her (Mamata Banerjee). We have shared a very old relationship with Didi, and that relationship will always remain the same. The respect we have for her will always remain. But one thing is true... everyone says that we received votes because of Didi. That is correct. We got votes because of Didi, because Didi herself is the identity and symbol of the Trinamool Congress. In Trinamool Congress, the face people recognise and trust is Didi’s face, and we all respect her deeply. People voted after seeing Didi, and they voted for Rachna Banerjee not because they were impressed by my face, but because they believed I would be able to work for them. Didi was the face behind it, and after that they gave us their votes with the expectation that I would do good work for the people and for my constituency. There was absolutely no resentment. But I felt that if the Central government and the State government were ruled by the same side, then getting work done would become much easier... something we have not really seen over the past 15 years." "People want us to do something for them, to carry out good work for them. And to make that happen, support from the Centre is very important. So at times, with due respect to Didi, I felt that while staying with her, there were many instances where obstacles were created in our work, and we were unable to do the kind of work we actually wanted to do.” #TMC #RachnaBanerjee (Full video available on PTI Videos -

Press Trust of India

197,146 views • 2 months ago

someone built a tool that puts ANY FACE on your webcam in REAL TIME called DOT - deepfake offensive toolkit, open source, works right now dot injects a real-time deepfake into your virtual camera zoom, google meet, teams, any video call app sees the fake face as your real webcam it ships with multiple deepfake engines (FOMM, SimSwap, SimSwapHQ), you pick the model based on your GPU and how real you need it to look built by sensity ai for red teams to test if your identity verification actually holds up against deepfakes

someone built a tool that puts ANY FACE on your webcam in REAL TIME called DOT - deepfake offensive toolkit, open source, works right now dot injects a real-time deepfake into your virtual camera zoom, google meet, teams, any video call app sees the fake face as your real webcam it ships with multiple deepfake engines (FOMM, SimSwap, SimSwapHQ), you pick the model based on your GPU and how real you need it to look built by sensity ai for red teams to test if your identity verification actually holds up against deepfakes

chiefofautism

37,088 views • 5 months ago

Yesterday I ran a full test inside INVIDEO with Nano Banana 2 (Nano Banana Pro). Invideo made it free and unlimited for 365 days on any paid plan till Nov 27, so the timing lined up and I pushed it through my full workflow. It blew me away. Here are the test cases I ran (the thread shows the results): - Realistic skin texture - Face swap and character replacement - Combine 4 images at a time with high-quality results - Multi-angle generation and Multi-profile character shots - Text adherence - Prompt adherence - Relighting Comment ‘invideo’ to get access. Now here’s each use case with RESULTS Like, REPOST and leave your thoughts.

Beginnersblog

18,559 views • 8 months ago

Scammers have been exposed for collecting money online under the guise of supporting Magyar and the Birds. They now face up to eight years in prison. A reminder: the Birds do not conduct any fundraising campaigns! However, for those who wish to contribute to our fight, verified donation options are available on the official Birds website (linked in our profile header) or through the United24 app, where you can also set up regular donations. The Birds are not on TikTok. Thanks to the National Police for their work. Video from Magyar – Ukrainian.

414 Magyar's Birds

88,043 views • 9 months ago

This is crazy! With LTX 2 in ComfyUI i was able to keep my original camera movement + it did a full lip-sync on the singing haha. All running locally. I made the original video back in 2021 when Metahumans first came out and we were working in Unreal Engine 4. This music video was some kind of result from my first experiments with Livelink and Control Rig at that time. Now the new AI improved version is not perfect and i didn’t train any lora for character consistency yet. I also didn’t have the original video file anymore so I downloaded the Youtube version as input haha. However the camera result + facial expressions are really promising! I wonder how it would handle more complex and fast shots though.

This is crazy! With LTX 2 in ComfyUI i was able to keep my original camera movement + it did a full lip-sync on the singing haha. All running locally. I made the original video back in 2021 when Metahumans first came out and we were working in Unreal Engine 4. This music video was some kind of result from my first experiments with Livelink and Control Rig at that time. Now the new AI improved version is not perfect and i didn’t train any lora for character consistency yet. I also didn’t have the original video file anymore so I downloaded the Youtube version as input haha. However the camera result + facial expressions are really promising! I wonder how it would handle more complex and fast shots though.

Jerome | InsaneUnreal

41,375 views • 6 months ago

Streetopia - Get Out Of Reet 🎬 Introducing the initial installment of an evolving tale brought to life by Generative AI. This captivating narrative has emerged from numerous poorman experiments and showcases the potential for an ongoing series of bite-sized episodes, each lasting under a minute. How I did do it? 1. All 3-4 shots of me are on green screen. 2. Backgrounds were created in a photo editor using stock pictures. Mostly stitched. Different aspect ratios. 3. I downloaded static 3D models & Animated models (greenscreen/alpha) and played with them in my motion graphics editor. Filled in some storefront windows with more videos. 4. Did some masking magic, camera movement, and played with layers and depth... for example, I masked the sky to add a storm and t-rex 🦖 5. Sent individual clips to Superstudio (in 4k) 6. Exported favorite renders (in 4k) and sent them to my video editor for choppin', sound design (poppin'), and voiceover (in 4k). See thread for examples. Challenges: • I wish I had a treadmill to actually look like I'm walking. • I was lucky with frame consistency. Except for the wide shot that fuzzed my face (me on the couch). I re-did the shot on Kaiber in a close-up and manually key-framed my face on that wide shot. • I couldn't do much about the buggy eyes in the last clip. • No butterflies were harmed in this video. How long did it take: • A day and a half with tons of happy accidents • How long will it take you? Probably much longer if you don't know how to shoot, edit, design, composite, edit sound, or know where to look for assets. But it's not impossible! What's the plan? To continue the story while still learning and experimenting along the way. This is a much longer process. Very tiring. Might try doing it weekly or bi-weekly. What's the story about? In the quest to escape Reet, everyone yearns for freedom. Ricardo, however, carries a hint of cynicism. He finds pleasure in the disorder and may seem reluctant to lend a hand. Nevertheless, the encounters with diverse characters along his odyssey will ignite a transformation within him, although not without challenges. Gradually, he will come to realize the profound truth: "Helping others is away is a way of helping ourselves." HQ Video: #GenerativeAI #ArtistOnTwitter #aiartcommunity #AIArtistCommunity #video2video

Streetopia - Get Out Of Reet 🎬 Introducing the initial installment of an evolving tale brought to life by Generative AI. This captivating narrative has emerged from numerous poorman experiments and showcases the potential for an ongoing series of bite-sized episodes, each lasting under a minute. How I did do it? 1. All 3-4 shots of me are on green screen. 2. Backgrounds were created in a photo editor using stock pictures. Mostly stitched. Different aspect ratios. 3. I downloaded static 3D models & Animated models (greenscreen/alpha) and played with them in my motion graphics editor. Filled in some storefront windows with more videos. 4. Did some masking magic, camera movement, and played with layers and depth... for example, I masked the sky to add a storm and t-rex 🦖 5. Sent individual clips to Superstudio (in 4k) 6. Exported favorite renders (in 4k) and sent them to my video editor for choppin', sound design (poppin'), and voiceover (in 4k). See thread for examples. Challenges: • I wish I had a treadmill to actually look like I'm walking. • I was lucky with frame consistency. Except for the wide shot that fuzzed my face (me on the couch). I re-did the shot on Kaiber in a close-up and manually key-framed my face on that wide shot. • I couldn't do much about the buggy eyes in the last clip. • No butterflies were harmed in this video. How long did it take: • A day and a half with tons of happy accidents • How long will it take you? Probably much longer if you don't know how to shoot, edit, design, composite, edit sound, or know where to look for assets. But it's not impossible! What's the plan? To continue the story while still learning and experimenting along the way. This is a much longer process. Very tiring. Might try doing it weekly or bi-weekly. What's the story about? In the quest to escape Reet, everyone yearns for freedom. Ricardo, however, carries a hint of cynicism. He finds pleasure in the disorder and may seem reluctant to lend a hand. Nevertheless, the encounters with diverse characters along his odyssey will ignite a transformation within him, although not without challenges. Gradually, he will come to realize the profound truth: "Helping others is away is a way of helping ourselves." HQ Video: #GenerativeAI #ArtistOnTwitter #aiartcommunity #AIArtistCommunity #video2video

Sway Molina

49,677 views • 3 years ago

This will be a multi-million $$ AI business: Deepfaked done-for-you social media content for busy founders and brands. This video is a 100% AI clone of my face and voice. Here's how it could work: One of the biggest headaches founders face is creating social media content. You know you need to do it, but: 1. Hiring a good content creator is like finding a needle in a haystack 2. Even if you find someone good, you realize the face of your brand should be you (founder-led) 3. Building a following on someone else's face is risky - they could leave any second So you realize the only real option is for the founder/CEO to create the content. BUT founders/CEOs obviously don't have time for that. The solution: hire a 'Ghostvideo' agency. → Clone founder's face and voice (with permission) → Agency creates and posts content, and knows how to analyze + grow accounts → Brand grows faster with "authentic" founder-led content → Founder focuses on actually running the company Benefits: 1. Social media accounts grow faster by having an agency solely focused on growing your account 2. Founder-led content (without the time suck) 3. Agency is replaceable, but the face of a brand is not Over the next 30 days, I'm testing this on my personal IG account (for those who want to follow along, my IG handle is the same as my Twitter handle). Fun fact: There's a major course platform and a massive Indian content creator that's already crushing it with this deepfake AI clone-style video content. If Ghostwriting agencies do north of $5-10m+ per year helping founders do this exact thing for Twitter and LI, done-for-you video will get even bigger. Video has been restricted for these agencies… until now with deepfaked AI clones. 'Ghostvideos' will be massive, and this demo my team and I created, while already pretty good, is the worst it'll ever be. P.S. If you're interested in learning how we produced this (with 3 separate AI tools) or want to work with us to create a "Ghostvideo" agency, DMs are open.

Rowan Cheung

323,341 views • 2 years ago