Anything World

Teachable Machine is an online tool that allows users to swiftly and effortlessly build machine learning models without the need for coding. It offers flexibility by enabling users to utilize files or capture examples in real time. This tool can classify images, audio, and poses, and the models generated can be exported for implementation in websites, applications, and beyond.

Omniverse Audio2Face is a reference application that facilitates the animation of a 3D character to align with any voice-over track. It comes equipped with a 3D character model that can be animated using a voice-over track, where the audio input is processed by a pre-trained Deep Neural Network. The output then manipulates the 3D vertices of the character mesh to generate real-time facial animation. Additionally, it includes a character transfer feature, enabling users to retarget any 3D human or human-like face and scale the output with multiple characters in the scene, while controlling the character's emotion through an AI network. It also offers data conversion features for blendshape conversion and blendweight export options, and it supports export-import with Blendshapes for Blender and Epic Games Unreal Engine to produce character motion.

Motorica focuses on cutting-edge character animation using generative deep learning techniques. Their leading product, MoGen, enables users to quickly produce top-notch locomotion by blending more than 150 styles. MoGen provides customizable motion paths and allows users to design their own within fbx-compatible software. Motorica's technology utilizes AI for innovative animation and has been awarded a MegaGrant from Epic Games for incorporating their AI-driven animation technology into the Unreal platform. This startup, set to revolutionize the conventional animation industry, is expected to release its first product in the first quarter of 2023.