AI actors, 3D printable robots, new AI image editors, AI music composer, 4D videos
Welcome to the AI Search newsletter. Here are the top highlights in AI this week.
DreamO image editor is an open-source tool that lets you edit and generate images using text prompts. You can describe what you want to see, and DreamO will create or modify images based on your instructions, making creative editing much easier. The project provides code and models for anyone to try or build on. Read more
Hunyuan Custom is an AI tool that generates personalized videos using text, images, audio, and video as input, keeping the subject’s identity consistent throughout. It fuses multiple input types and uses advanced modules to ensure the person or object in the video always looks the same, even as they move or change expressions. This open-source model is designed for developers and businesses to create high-quality, controlled video content. Read more
PixelHacker is a free AI tool that erases or fills in missing parts of images with realistic details. It uses AI to understand both the structure and meaning of the image, ensuring seamless edits. Read more
Do you prefer to watch instead of read? Check out this video covering all the highlights in AI this week:
Researchers have developed a system that lets robots identify an object's properties through handling. The system uses a simulation process that incorporates models of the robot and the object to rapidly identify characteristics of the object as the robot interacts with it. This low-cost technique could be especially useful in applications where cameras might be less effective. Read more
A new mathematical model predicts how well a neural network can learn from limited data. This model combines two analytical methods to accurately assess how well a neural network can generalize data when it adopts knowledge from another network. This research can improve AI training in fields where data is scarce, such as medical diagnostics. Read more
We’re partnering with Dell to give away a Dell Precision 5690 Workstation with a RTX 5000 Ada. This is a powerful yet portable laptop, with a built-in NPU that’s optimized for AI. Only available to USA or Canada residents. Enter for FREE here
ACE-Step is a free, open-source AI music generator that creates high-quality songs super fast, supporting up to 19 languages. It lets you control details like voice cloning, lyric editing, and remixing, and can make a 4-minute track in just 20 seconds on powerful hardware. Artists, producers, and developers can use it for everything from songwriting to making instrumentals and vocal samples. Read more
Researchers at MIT have developed a ping-pong-playing robot that can return shots with high-speed precision. The robot uses a combination of high-speed cameras and predictive control to estimate the speed and trajectory of an incoming ball and execute a precise swing. Its strike speed approaches that of human players, making it a potential competitor in robotic table tennis. Read more
Amuse that is an AI can assist music composers by transforming text, images, or audio inspirations into chord progressions. Amuse uses a combination of a large language model and a filtering model trained on real music data to generate suggestions that respect the user's creative flow. This allows for flexible and user-centered collaboration between the composer and the AI system. Read more
Berkeley Humanoid Lite is an open-source humanoid robot that anyone can build for under $5,000 using a standard 3D printer and off-the-shelf parts. The robot stands about 0.8 meters tall, weighs 16 kg, and features modular 3D-printed gearboxes, making it affordable, customizable, and easy to repair or modify. All hardware designs, code, and training resources are freely available, aiming to make advanced robotics accessible to students, researchers, and hobbyists worldwide. Read more
LTX Video 13B is a new AI video model by Lightricks that creates high-quality videos super fast, even on regular computers. It uses a “multiscale rendering” process to build videos in layers, making the process over 30 times faster than similar models while keeping great detail and realism. The model is open-source and uses advanced compression to make video creation more accessible to everyone. Read more
With Monica, you can use the top AI models, image generators, and video generators, all in one integrated platform. Use code AISEARCH10 to get 25% OFF 'Unlimited Annual Plan' within 24h of registration, or enjoy 10% OFF. Try it for free today!
FlexiAct transfers actions from one video to a different subject image. It adapts movements across different body types and angles while keeping the subject's identity intact. Read more
HoloTime 4D scenes is a framework that turns a single panoramic image into a fully immersive 360-degree 4D scene. It uses AI to animate the image into a panoramic video, then reconstructs it into a 4D environment that you can explore in VR or AR. This makes it possible to create realistic, interactive virtual worlds from just one image. Read more