In nuce: At the CES 2023 (Consumer Electronics Show), graphics card manufacturer Nvidia announced various innovations for its Omniverse collaboration platform. This is according to an announcement from 3 January 2023, published by Omniverse Vice President Frank DeLise on blogs.nvidia.com. The announced highlights include: AI tools for 3D artists, an extension for Blender and the early access of RTX Remixer (for the purpose of ray-tracing and DLSS-based remasters of gaming classics).

In toto: Below is an outline of the announced highlights for Omniverse – and the corresponding video of Nvidia’s special address at CES 2023, moderated by Jeff Fisher, Senior Vice President at Nvidia.

NVIDIA Special Address at CES 2023

• New, experimental AI tools for 3D artists: This concerns the tools Audio2Face, Audio2Gesture and Audio2Emotion. These are tools for 3D character animation, all of which have received a performance update.
• Extension for Blender: The Blender Alpha version is now available for the Omniverse Launcher. This allows geometries to be repaired, UVs to be generated automatically and high-resolution CAD files (computer-aided design) to be downsampled to a more practical polycount. Users can now easily transfer shape keys and rigged characters – or attach the head of one from Omniverse Audio2Face. Audio2Face is an AI-powered tool that automatically creates realistic facial expressions based on audio files.
• Nvidia RTX Remix will soon be available as Early Access: Such classic games that caused a sensation with RTX remasters from the modder community include Half-Life 2 and Max Payne. RTX Remix is a modding platform that – as stated on nvidia.com – utilises “powerful AI tools” to “enhance classic games and quickly create stunning RTX remasters with ray tracing and DLSS (Deep Learning Super Sampling) .”
• Update for Canvas: This will be available to users of Nvidia’s RTX graphics cards. Canvas is an AI tool that can be used to create low-threshold landscapes.
• Omniverse will soon be pre-installed on the first Nvidia Studio laptops: As Nvidia mentioned, laptops from Taiwanese computer hardware manufacturer Gigabyte Technology will already ship with the Omniverse Launcher this year (this also includes laptops from Gigabyte’s gaming division Aorus)
• New, free USD assets: The number of these should be in the thousands.

Turn the page: In our previous three-part advertorial series on Nvidia’s Omniverse, we reported on the basics of the collaboration platform (“Omniverse” on page 56 of DP 03 : 2022), how it can be set up professionally (“Getting started with Omniverse” on p. 84 of DP 04 : 2022) and how a sensible workflow can be set up (“Working in Omniverse” on p. 87 of DP 05 : 2022; the issue is also available as a free trial copy!).

Clicked further: Nvidia’s announcement, published in the course of CES 2023, can be viewed at blogs.nvidia.com . On 9 January 2023 , Digital Production recently reported on Omniverse – at that time, an extensive update for Omniverse Enterprise was worth reporting.

Sources: 80.lv ( announcement by Gloria Levine), blogs.nvidia.com ( announcement by Frank Delise)

The post Omniverse updates at the CES 2023! first appeared on DIGITAL PRODUCTION and was written by Patrick Poti.

In nuce: The US company OpenAI, which researches artificial intelligence, has published Point-E. Point-E can be used to create 3D objects using text input. This is according to a report by journalist Andrew Tarantola, published on the Engadget news service. Point-E works in a similar way to another development from OpenAI: Dall-E, the text-image generator. The difference is that Point-E creates 3D models – Dall-E only creates still images. The background to this: Dall-E is an artificial intelligence that is used to generate images based on text input (these text inputs are also known as prompts in technical jargon). Digital Production last reported on Dall-E on 27 October 2022 , when the generator was integrated into the Microsoft Bing internet search engine.

In toto: Point-E is used to create 3D point clouds, also based on prompts. While DreamFusion from Google – as Andrew Tarantola writes – “usually takes several hours to create 3D objects”, Point-E minimises the process to just a few minutes. In addition, Tarantola goes on to say that Point-E should already work properly with just one GPU – instead of, as with DreamFusion, only producing results when several GPUs are used.

How does Point-E work? In the scientific paper Point-E: A System for Generating 3D Point Clouds from Complex Prompt , the team led by Alex Nichol, researcher at OpenAI, writes that Point-E utilises “a large corpus of (text, image) pairs”. However, Nichol and his team go on to explain that Point-E was trained using a “smaller dataset of (image, 3D) pairs”. The paper also discusses how 3D objects are created with the support of Point-E. The article states that in the case of a text input (e.g. “A cat eats a burrito”), Point-E first generates a synthetic 3D representation of a burrito-eating cat. The image generated in this way is then run through several diffusion models; these models serve as the basis for a 3D RGB point cloud – first for a coarse 1,024-point cloud model, then a finer 4,096-point model.

Click to continue: on 6 September 2022 , Digital Production showed how the text-to-image AI Stable Diffusion allows users to create images using text input.

Source: engadget.com ( report by Andrew Tarantola)

The post Enter text – and a finished 3D object appears? first appeared on DIGITAL PRODUCTION and was written by Patrick Poti.

In nuce: The technology company Rokoko has released free, AI-supported software: Rokoko Video. This software, supported by artificial intelligence, is tailored to motion capturing – and can be controlled via the web browser. Rokoko announced the release of Rokoko Video on 6 December 2022 via the website 3dvf.com .

Start animating with our free AI mocap tool Rokoko Video I Tutorial

In toto: According to the announcement on the company website, Rokoko Video is comparatively easy to use. The first step is for users to upload a video using a webcam, smartphone or comparable moving image recording device. It should be noted that the entire body should be recorded, which users can then edit using motion capturing. The motion data is generated automatically by the software. If required, the same data can be edited using Rokoko Studio (also from Rokoko)(Manuel Skroblin reported in detail for Digital Production on 1 March 2019 about how character animation can look with Rokoko Studio and associated hardware and software solutions). Finally, users import the finished motion capturing into common DCC software such as Maya, Blender, Cinema 4D or Unreal Engine. According to the company, a free version of Rokoko Studio is also available.

Click further: Further information is available via the aforementioned website at 3dvf.com . An alternative to Rokoko Video is Maxon Moves, an app for mobile devices, which Digital Production reported on 5 March 2020 .

The post Motion capturing with Rokoko Video? first appeared on DIGITAL PRODUCTION and was written by Patrick Poti.

In nuce: After more than ten years of development, the full version of the AI-supported keyframe animation software Cascadeur will be released within the next few days. In the course of this, a five-minute video has been published which presents the functions of the full version. This is based on information available to the editorial team.

In toto: The release marks the end of an almost three-year beta testing phase in which more than 100,000 industry professionals and hobbyists were involved. The video is linked below for those interested.

Click further: Digital Production last reported on Cascadeur on 20 August last year , when the software went into early access. In our print and e-paper issue DP 02 : 2021 , our author Ralf Gliffe took a closer look at the Cascadeur open beta on page 6.

Cascadeur release version product demo

The post The Cascadeur animation software is available as a full version! first appeared on DIGITAL PRODUCTION and was written by Patrick Poti.

In nuce: Russell Midfield, a Blender fanboy with a focus on creating characters, has published a ten-minute video in which he compares four motion capture programmes. The highlight: all four programmes work on the basis of artificial intelligence.

In toto: For his video, Midfield takes a look at the programmes Plask, Radical, DeepMotion and Pixcap (Plask was presented by Digital Production on 26 January this year ) – and uses iClone 8 to solve some of the typical teething troubles of motion capture software, such as posture correction, shaky movements and slipping feet.

Click further: Digital Production reported on the release of iClone 8 on 7 June 2022. Russell Midfield’s entire video is linked below.

Russell Review: AI Motion Capture – Who is the best?

The post What can iClone 8 from Reallusion do that the competition can’t? first appeared on DIGITAL PRODUCTION and was written by Patrick Poti.