Humanoid Robots on a Typical Friday? Maybe in a few years. This week, I combined the latest open-source WAN tools from Alibaba Group with Gemini, added it with ComfyUI, and integrated my own workflows. The result: WAN’s new Animate tool is impressive. WAN 2.2 is essentially an upgraded version of 2.1, trained with 65.6% more images and 83.2% more videos. The Animate model, based on WAN 2.2, uses reference videos to generate pose animations. To get started, you need two things: a reference image and a reference video to guide the animation. Take your smartphone, have yourself filmed while playing with your dog, and transfer the poses onto your desired character. The improvements in WAN 2.2 mean smoother animations, more accurate poses, and faster workflow integration. Excited to see how these tools will continue to push creative possibilities in animation! #ai #comfyui #wan #commercials
More Relevant Posts
-
I burnt my precious AI credits trying to recreate a simple header animation. Still think AI can replace humans? Think again! Last week, I was working on my first YouTube video, trying to recreate an animation from MKBHD's Panels app. I gave it as much context as possible, extracted every frame of the animation, and wrote the most detailed prompt you've ever seen. But after countless back-and-forths, the animation still looked broken. Then I remembered a technique I saw in the Reanimated documentation the week before. It wasn’t exactly what I needed, but I understood the principle. I deleted everything the AI had generated and implemented the animation myself. 15 minutes. Done! The solution was surprisingly simple, but the AI kept overengineering it. Here's what I learned: AI can't take your job if you're a problem solver, not just a coder. Anyone can now code with AI, but it often needs experienced guidance. What do you think? Are we becoming too dependent on AI for problems we could actually solve faster ourselves?
To view or add a comment, sign in
-
The process of 𝐢𝐧-𝐛𝐞𝐭𝐰𝐞𝐞𝐧𝐢𝐧𝐠, the delicate art of drawing frames between key poses, is the final, most critical step in transforming rough sketches into a flowing, cinematic masterpiece. But even for the most seasoned animator, the "tweening phase" is a black box. Slight, undetectable shifts in timing, easing, and arc can derail a sequence, leading to motion inconsistency, jitters, and missed opportunities for the optimal quality and style profile. The challenge is to bring 𝐩𝐫𝐞𝐜𝐢𝐬𝐢𝐨𝐧 𝐬𝐜𝐢𝐞𝐧𝐜𝐞 𝐭𝐨 𝐭𝐡𝐢𝐬 𝐚𝐧𝐜𝐢𝐞𝐧𝐭 𝐜𝐫𝐚𝐟𝐭. We've developed an AI-Powered "𝐀𝐮𝐭𝐨𝐦𝐚𝐭𝐞𝐝 𝐀𝐧𝐢𝐦𝐚𝐭𝐢𝐨𝐧 𝐈𝐧-𝐁𝐞𝐭𝐰𝐞𝐞𝐧𝐢𝐧𝐠" Platform that acts as a digital tweening master. By continuously analyzing the key visual and timing indicators in your keyframes, the AI uses advanced machine learning to correlate these variables with historical animation data, allowing it to generate the final, smooth animation with unprecedented accuracy. This moves your process from intuition and hundreds of hours of manual work to objective data. 𝐀𝐮𝐭𝐨𝐦𝐚𝐭𝐞𝐝 𝐀𝐧𝐢𝐦𝐚𝐭𝐢𝐨𝐧 𝐈𝐧-𝐁𝐞𝐭𝐰𝐞𝐞𝐧𝐢𝐧𝐠: 𝐀 𝐆𝐮𝐢𝐝𝐞 𝐭𝐨 𝐈𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐭 𝐌𝐨𝐭𝐢𝐨𝐧 𝐚𝐧𝐝 𝐏𝐫𝐞𝐝𝐢𝐜𝐭𝐢𝐨𝐧 We provide a service that equips animators with the tools to master consistency and achieve their target motion profiles every single time. Stop guessing and start predicting. Utilize the power of AI to: 🔹 𝐀𝐧𝐚𝐥𝐲𝐳𝐞 keyframes, character models, and specified timing for motion intent. 🔹 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐞 all the "in-between" frames, creating smooth, high-quality animation automatically. 🔹 𝐑𝐞𝐜𝐨𝐦𝐦𝐞𝐧𝐝 the 𝐨𝐩𝐭𝐢𝐦𝐚𝐥 𝐞𝐚𝐬𝐢𝐧𝐠 𝐜𝐮𝐫𝐯𝐞𝐬 and 𝐚𝐫𝐜𝐬 to hit a precise style profile. 🔹 𝐆𝐮𝐚𝐫𝐚𝐧𝐭𝐞𝐞 motion quality and consistency across every sequence and shot. Upgrade your craft and engineer perfection in every motion. Learn more at 𝐡𝐭𝐭𝐩𝐬://𝐯𝐚𝐫𝐞𝐧𝐲𝐚𝐳.𝐜𝐨𝐦/ and contact us at 𝐜𝐨𝐟𝐟𝐞𝐞@𝐯𝐚𝐫𝐞𝐧𝐲𝐚𝐳.𝐜𝐨𝐦. #AnimationAutomation #AIinAnimation #TweeningTool #DigitalArt #2DAnimation #TechInnovation #Gamedev #Filmmaking
To view or add a comment, sign in
-
Midjourney + Kinetix → Character Transformation (here's the full process) 1. Generate a character image with Midjourney. 2. Record a short video (of yourself or anyone). 3. Upload the image and the video to Kinetix. 4. Kinetix turns your performance into animation. It's the first model with a 3D native approach. → The model processes your input video in the background, converting it into a 3D file. → This 3D file is the base for a new video, allowing you to select different camera moves and angles. It's a huge leap forward for AI videos. Just act it out, Kinetix generates it. PS: I test AI tools daily, and I update a list of my favorite ones. Click here to access it → best-ai.blog
To view or add a comment, sign in
-
From static images to dynamic 3D animations — in minutes. This Midjourney → Kinetix pipeline is more than a workflow. It’s a glimpse of how AI is transforming creativity. In the near future, anyone can be an actor, director, and animator of their own story.
Midjourney + Kinetix → Character Transformation (here's the full process) 1. Generate a character image with Midjourney. 2. Record a short video (of yourself or anyone). 3. Upload the image and the video to Kinetix. 4. Kinetix turns your performance into animation. It's the first model with a 3D native approach. → The model processes your input video in the background, converting it into a 3D file. → This 3D file is the base for a new video, allowing you to select different camera moves and angles. It's a huge leap forward for AI videos. Just act it out, Kinetix generates it. PS: I test AI tools daily, and I update a list of my favorite ones. Click here to access it → best-ai.blog
To view or add a comment, sign in
-
Concept AV: Robi 5G. Speed Unleashed! Here is a concept AV I created a month ago to imagine the launch of the country’s first Robi 5G. I wanted to experiment with Nano Banana to see if I could generate different iterations of Siam Ahmed across different scenes from a single image (reference image attached in comments). Honestly, it’s not perfect, yet, but it demonstrates the exciting potential of AI in shaping the future of advertising and storytelling. Tools used: Images: Nano Banana, Midjourney, ImageFX, Animation: Midjourney, Veo3, Kling AI Edit/Compositing: Photoshop, After Effects Huge thanks to Sabbir Hasan and Sonjoy Devnath for their valuable inputs. Would love to hear your thoughts on this!
To view or add a comment, sign in
-
This AI model creates ultra-realistic images with perfect consistency across poses, angles, and scenes. On the ImagineArt platform, you can not only generate images but also edit and animate them — using all the newest models. With NanoBanana, you can restyle outfits and backgrounds, and with models like Kling or Sora2, you can turn your creations into full animations. A single platform where creativity has no limits. ✨ #ImagineArt #ImagineArt10 #AllImageGenerator #RealisticAIImages #CharacterCreationAI #BrandingWithAI #BusinessAI #GoogleNanoBanana #ProductIntegrationAI #GenerativeAI #CreativeAI #AIForCreators #AIMagic #AIContentCreation #NextGenImagery #SmartAI #DigitalCreators #InnovativeAI #FutureOfBranding #CreatorEconomy
To view or add a comment, sign in
-
What makes Apple Intelligence and #Gemini AI #visuals feel so organic and alive? The secret lies in shaders - a technology that enables high-performance parallel pixel manipulation beyond the boundaries of traditional layouts and components. As companies shift from flat design to dynamic, organic animations, shaders are becoming essential for standing out. Florian Vögtle explores: → Fundamentals of shaders and their primary use cases → How to integrate shader-based experiences into modern apps → Practical examples: AI integrations, dynamic scrolling, interactive buttons You'll leave with everything needed to design and implement shader-based experiences that bring your app to life. #DevFestHamburg #Shaders #AppDevelopment #UI #Animation #MobileDev #TechCommunity
To view or add a comment, sign in
-
-
Stefan’s AI-generated Porsche spec ad took over 80 hours and $500 in credits to craft, combining tools like Midjourney, Photoshop, Magnific AI, Kling, and Veo, with manual animation, lighting, and sound design. Initially rejected on Reddit, the project later went viral after an AI-focused company recognized its skill, proving that even with AI, creativity and craftsmanship drive impact. #awesmai #AI #aigeneratedvideo
To view or add a comment, sign in
-
When it comes to video generation, there are endless possibilities of what you can do with your creative mind: Lets break it down: this video is purely AI from AI to Z from the images to the animations and the sound. Tools Used: ✅ Gemini AI for image generation (FREE) ✅ Vheer AI for animation, Basic Image Prompt and Camera Prompt (FREE) ✅ Capcut for car sound track (FREE but you can upgrade to pro) How I got my results: ✅I prompted Gemini AI to generate image of a sports car ✅I imported the image i generated from Gemini AI to Vheer and added the basic image prompt plus camera prompt and click on generate an animated image ✅I imported the animated clip to Capcut and added the car race sound track just to spice it up a bit and make it more realistic The output is what you have on your screen. Pls like, comment and share but not to forget your questions are also important Florence Chinenye Uguba
To view or add a comment, sign in
-
A perspective of creativity turned passion for the love of limitless possibility of AI Lets break it down: From A-Z is AI oriented that is from the images to the animations and the sound. Tools Used: ✅ Gemini AI for image generation (FREE) ✅ Vheer AI for animation, Basic Image Prompt and Camera Prompt (FREE) ✅ Pixabay for sound track ✅ Capcut for editing and compilation of clips (FREE but you can upgrade to pro) How I got my results: ✅I prompted Gemini AI to generate image of hyper realistic future car ✅I imported the image i generated from Gemini AI to Vheer and added the basic image prompt plus camera prompt and click on generate an animated image ✅ I download the sound track from Pixabay ✅I imported the animated clips to Capcut and added the downloaded sound track from Pixabay The output is what you have on your screen. Pls like, comment and share but not to forget your questions are also important
To view or add a comment, sign in