Partenariat Le freelancing est une voie naturelle pour nombre d’artistes et techniciens de la 3D, qu’ils viennent de l’animation, du jeu vidéo, de l’architecture ou des effets visuels. En parallèle d’une explosion des besoins en contenus visuels temp

Generative AI has reshaped how people create, imagine and interact with digital content. As AI models continue to grow in capability and complexity, they require more VRAM, or video random access memory. The base Stable Diffusion 3.5 Large model, for example, uses over 18GB of VRAM — limiting the number of systems that can run it well. By applying quantization to the model, noncritical layers can be removed or run with lower precision. NVIDIA GeForce RTX 40 Series and the Ada Lovelace generation of NVIDIA RTX PRO GPUs support FP8 quantization to help run these quantized models, and the latest-generation NVIDIA Blackwell GPUs also add support for FP4. NVIDIA collaborated with Stability AI to quantize its latest model, Stable Diffusion (SD) 3.5 Large, to FP8 — reducing VRAM consumption by 40%. Further optimizations to SD3.5 Large and Medium with the NVIDIA TensorRT software development kit (SDK) double performance. In addition, TensorRT has been reimagined for RTX AI PCs, combining its industry-leading performance with just-in-time (JIT), on-device engine building and an 8x smaller package size for seamless AI deployment to more than 100 million RTX AI PCs. TensorRT for RTX is now available as a standalone SDK for developers. RTX-Accelerated AI NVIDIA and Stability AI are boosting the performance and reducing the VRAM requirements of Stable Diffusion 3.5, one of the world’s most popular AI image models. With NVIDIA TensorRT acceleration and quantization, users can now generate and edit images faster and more efficiently on NVIDIA RTX GPUs. Stable Diffusion 3.5 quantized FP8 (right) generates images in half the time with similar quality as FP16 (left). Prompt: A serene mountain lake at sunrise, crystal clear water reflecting snow-capped peaks, lush pine trees along the shore, soft morning mist, photorealistic, vibrant colors, high resolution. To address the VRAM limitations of SD3.5 Large, the model was quantized with TensorRT to FP8, reducing the VRAM requirement by 40% to 11GB. This means five GeForce RTX 50 Series GPUs can run the model from memory instead of just one. SD3.5 Large and Medium models were also optimized with TensorRT, an AI backend for taking full advantage of Tensor Cores. TensorRT optimizes a model’s weights and graph — the instructions on how to run a model — specifically for RTX GPUs. FP8 TensorRT boosts SD3.5 Large performance by 2.3x vs. BF16 PyTorch, with 40% less memory use. For SD3.5 Medium, BF16 TensorRT delivers a 1.7x speedup. Combined, FP8 TensorRT delivers a 2.3x performance boost on SD3.5 Large compared with running the original models in BF16 PyTorch, while using 40% less memory. And in SD3.5 Medium, BF16 TensorRT provides a 1.7x performance increase compared with BF16 PyTorch. The optimized models are now available on Stability AI’s Hugging Face page. NVIDIA and Stability AI are also collaborating to release SD3.5 as an NVIDIA NIM microservice, making it easier for creators and developers to access and deploy the model for a wide range of applications. The NIM microservice is expected to be released in July. TensorRT for RTX SDK Released Announced at Microsoft Build — and already available as part of the new Windows ML framework in preview — TensorRT for RTX is now available as a standalone SDK for developers. Previously, developers needed to pre-generate and package TensorRT engines for each class of GPU — a process that would yield GPU-specific optimizations but required significant time. With the new version of TensorRT, developers can create a generic TensorRT engine that’s optimized on device in seconds. This JIT compilation approach can be done in the background during installation or when they first use the feature. The easy-to-integrate SDK is now 8x smaller and can be invoked through Windows ML — Microsoft’s new AI inference backend in Windows. Developers can download the new standalone SDK from the NVIDIA Developer page or test it in the Windows ML preview. For more details, read this NVIDIA technical blog and this Microsoft Build recap. Join NVIDIA at GTC Paris At NVIDIA GTC Paris at VivaTech — Europe’s biggest startup and tech event — NVIDIA founder and CEO Jensen Huang yesterday delivered a keynote address on the latest breakthroughs in cloud AI infrastructure, agentic AI and physical AI. Watch a replay. GTC Paris runs through Thursday, June 12, with hands-on demos and sessions led by industry leaders. Whether attending in person or joining online, there’s still plenty to explore at the event. Each week, the RTX AI Garage blog series features community-driven AI innovations and content for those looking to learn more about NVIDIA NIM microservices and AI Blueprints, as well as building AI agents, creative workflows, digital humans, productivity apps and more on AI PCs and workstations.  Plug in to NVIDIA AI PC on Facebook, Instagram, TikTok and X — and stay informed by subscribing to the RTX AI PC newsletter. Follow NVIDIA Workstation on LinkedIn and X.  See notice regarding software product information.

🚀 Get started with the V-Ray ArchViz Collection → https://bit.ly/AnimateFaster V-Ray delivers world-class renders. But when your projects call for animated people, camera movement, or fast client feedback—traditional workflows can slow you down. In this video, we’ll show you how to combine V-Ray, Anima, and Chaos Vantage to create dynamic, animated scenes—and explore them in real time. --------------------------------------------------------------------------- Imagine. Design. Believe. Chaos provides world-class visualization solutions helping you share ideas, optimize workflows and create immersive experiences. Everything you need to visualize your ideas from start to finish. From architecture and VFX to product design and e-commerce, Chaos empowers creators to bring their projects to life. Our industry-leading tools, including V-Ray, Enscape, and Corona, are built for architects, designers, AEC professionals, and CG artists. Whether you’re crafting photorealistic visuals, immersive real-time experiences, or cinematic VFX, Chaos delivers the power and flexibility to render anything. 🔎 Explore Chaos products → https://bit.ly/ExploreChaos 🎓 Learn more & get free tutorials → https://bit.ly/ChaosWebinars 🔔 Subscribe for the latest updates! 🔗 Follow us: 👉 LinkedIn: https://bit.ly/ChaosLinkedIn 👉 Instagram: https://bit.ly/ChaosIG 👉 Facebook: https://bit.ly/Chaos_Facebook #Chaos #V-Ray #3DRendering #Visualization

In this Q&A, get a glimpse into the future of artificial intelligence (AI) and computer vision through the lens of longtime Unity user Gerard Espona, whose robot digital twin project was featured in the Made with Unity: AI series. Working as simulation lead at Luxonis, whose core technology makes it possible to embed human-level perception into robotics, Espona uses his years of experience in the industry to weigh in on the current state and anticipated progression of computer vision.During recent years, computer vision (CV) and AI have become the fastest-growing fields both in market size and industry adoption rate. Spatial CV and edge AI have been used to improve and automate repetitive tasks as well as complex processes.This new reality is thanks to the democratization of CV/AI. Increasingly affordable hardware access, including depth perception capability as well as improvements in machine learning (ML), has enabled the deployment of real solutions on edge CV/AI systems.Spatial CV using edge AI enables depth-based applications to be deployed without the need of a data center service, and also allows the user to preserve privacy by processing images on the device itself.Along with more accessible hardware, software and machine learning workflows are undergoing important improvements. Although they are still very specialized and full of technical challenges, they have become much more accessible, offering tools that allow users to train their own models.Within the standard ML pipeline/workflow, large-scale edge computing and deployment can still pose issues. One of the biggest general challenges is to reduce the costs and timelines currently required to create and/or improve machine learning models on real-world applications. In other words, the challenge is how to manage all these devices to enable a smooth pipeline for continuous improvement.Also, the implicit limitations in terms of compute processing need extra effort on the final model deployed on the device (that is, apps need to be lightweight, performant, etc.). That said, embedded technology evolves really fast, and each iteration is a big leap in processing capabilities.Spatial CV/AI is a field that still requires a lot of specialization and systems. Workflows are often complicated and tedious due to numerous technical challenges, so a lot of time is devoted to smoothing out the workflow instead of focusing on value-added tasks.Creating datasets (collecting and filtering images and videos), annotating the images, preprocessing/augmentation process, training, deploying and closing the feedback loop for continuous improvement is a complex process. Each step of the workflow is technically difficult and usually involves time and financial cost, and more so for systems working in remote areas with limited connectivity.At Luxonis, we help our customers build and deploy solutions to solve and automate complex tasks at scale, so we’re facing all these issues directly. Our mission, “Robotic vision made simple,” provides not only great and affordable depth-capable hardware, but also a solid and smooth ML pipeline with synthetic datasets and simulation.Another important challenge is the work that needs to be done on the interpretability of models and the creation of datasets from an ethical, privacy and bias point of view.Last but not least, global chip supply issues are making it difficult to get the hardware into everybody’s hands.Data-centric AI is potentially useful when a working model is underperforming. Investing a large amount of time to optimize the model often leads to almost zero real improvement. Instead, with data-centric AI the investment is in analysis, cleaning and improving of the dataset.Usually when a model is underperforming, the issue is within the dataset itself, as there is not enough data for the model to outperform. This could be the result of two possible reasons: 1) the model needs a much larger amount of data, which is difficult to collect in the real world, or 2) the model doesn’t have enough examples of rare cases, which take a lot of time to happen in the real world.In both situations, synthetic datasets could help.Thanks to Unity’s computer vision tools, it is very easy to create photorealistic scenes and randomize elements like materials, light conditions and object placement. The tools come with common labels like 2D bounding boxes, 3D bounding boxes, semantic and instance segmentation, and even human body key points. Additionally, these can be easily extended with custom randomizers, labelers and annotations.Almost any task you want to automate or improve using edge CV/AI very likely involves detecting people for obvious safety and security reasons. It’s critical to guarantee user safety around autonomous systems or robots when they’re working, requiring models to be trained on data about humans.That means we need to capture a large amount of images, including information like poses and physical appearance, that are representative of the entire human population. This task raises some concerns about privacy, ethics and bias when starting to capture real human data to train the model.Fortunately, we can use synthetic datasets to mitigate some of these concerns using human 3D models and poses. A very good example is the work done by the Unity team with PeopleSansPeople.PeopleSansPeople is a human-centric synthetic dataset creator using 3D models and standard animations to randomize human body poses. Also, we can use a Unity project template, to which we add our own 3D models and poses to create our own human synthetic dataset.At Luxonis, we’re using this project as the basis for creating our own human synthetic dataset and training models. In general, we use Unity’s computer vision tools to create large and complex datasets with a high level of customization on labelers, annotations and randomizations. This allows our ML team to iterate faster with our customers, without needing to wait for real-world data collection and manual annotation.Since the introduction of transformer architecture, CV tasks are more accessible. Generative models like DALL-E 2 could also be used to create synthetic datasets, and NeRF as a neural approach to generate novel point of views of known objects and scenes. It’s clear all these innovations are catching the attention of audiences.On the other hand, having access to better annotation tools and model zoos and libraries with pre-trained, ready-to-use models are helping drive wide adoption.One key element contributing to the uptick in computer vision use is the fast evolution of vision processing units (VPUs) that currently allow users to perform model inferences on device (without the need for any host) at 4 TOPS of processing power (current Intel Movidius Myriad X). The new generation of VPUs promises a big leap in capabilities, allowing even more complex CV/AI applications to be deployed on edge.Any application related to agriculture and farming always captures my attention. For example, there is now a cow tracking and monitoring CV/AI application using drones.Our thanks to Gerard for sharing his perspective with us – keep up with his latest thoughts on LinkedIn and Twitter. And, learn more about how Unity can help your team generate synthetic data to improve computer vision model training with Unity Computer Vision.

In this video, Martin Klekner talks about why he stepped away from the semi-realistic style of his previous Heroes of Bronze short film, to pursue the NPR (Non-photoreammo listic) style and finally learn Grease Pencil :-) Source

Valve's Steam Deck is getting a shot in the arm via NVIDIA, with the GPU manufacturer's GeForce Now streaming service arriving on the handheld PC – here's all you need to knowTech15:05, 29 May 2025Steam Deck will gain a huge library of titlesWe've already explained how we love the Steam Deck for PC gaming on the go, and with more and more amazing games in 2025 hitting Valve's platform like Assassin's Creed: Shadows and the ludicrously addictive Monster Train 2, it's not slowing down soon.Now, NVIDIA has confirmed that its GeForce Now streaming service, which gives users access to a powerful PC via the cloud, will offer a native app on the platform.‌"Members will be able to play over 2,100 titles from the GeForce NOW cloud library at GeForce RTX quality on Valve’s popular Steam Deck device with the launch of a native GeForce NOW app, coming later this year," NVIDIA explained.‌"Steam Deck gamers can gain access to all the same benefits as GeForce RTX 4080 GPU owners with a GeForce NOW Ultimate membership, including NVIDIA DLSS 3 technology for the highest frame rates and NVIDIA Reflex for ultra-low latency."NVIDIA says battery life will be improved while streaming games than playing natively, and will be ideal for playing docked, too.‌"The streaming experience with GeForce NOW looks stunning, whichever way Steam Deck users want to play — whether that’s in handheld mode for HDR-quality graphics, connected to a monitor for up to 1440p 120 fps HDR or hooked up to a TV for big-screen streaming at up to 4K 60 HDR," the blog post explains."GeForce NOW members can take advantage of RTX ON with the Steam Deck for photorealistic gameplay on supported titles, as well as HDR10 and SDR10 when connected to a compatible display for richer, more accurate colour gradients."NVIDIA adding a dedicated app on the platform for its service opens up questions about whether other services could do the same.‌Microsoft is reportedly working on a handheld PC in the vein of the Steam Deck, and fans have long asked for Game Pass functionality on Valve's storefront.Could NVIDIA open the door to an Xbox streaming service? Time will tell.It's not just Steam Deck, either, with GeForce NOW planned for Apple Vision Pro, Pico headsets, and Meta Quest 3 and 3S.Article continues below"Later this month, these supported devices will give members access to an extensive library of games to stream through GeForce NOW by opening the browser to play.geforcenow.com when the newest app update, version 2.0.70, starts rolling out later this month," NVIDIA explains."Members can transform the space around them into a personal gaming theatre with GeForce NOW. The streaming experience on these devices will support gamepad-compatible titles for members to play their favourite PC games on a massive virtual screen."For the latest breaking news and stories from across the globe from the Daily Star, sign up for our newsletters.‌‌‌

Got a project that’s too bold to build? Submit your conceptual works, images and ideas for global recognition and print publication in the 2025 Vision Awards! The Main Entry deadline of June 6th is fast approach — submit your work today. Architectural visualization has gone from a technical exercise to a creative discipline in its own right. Once treated as a (more or less) behind-the-scenes tool for client approval, rendering is now front and center, circulating online, shaping public perception, and winning awards of its own. There are many reasons for this shift. More powerful software, changing client expectations, and a deeper understanding of what visualizations can actually do have all contributed to it. As a result, photorealism has definitely reached staggering levels of clarity, but that’s just one part of the story. In this new era of rendering, visualizations also have a role in exploring what a building or a space could represent, evoke or question. This conceit is precisely why the Architizer’s Vision Awards were created. With categories for every style and approach, the program highlights the artists, studios and images pushing architectural rendering forward. With that in mind, take a closer look at what defines this new era and explore the Vision Awards categories we’ve selected to help you find where your work belongs. Rendering Is Now Part of the Design Process New Smyril Line headquarters, Tórshavn, Faroe Islands by ELEMENT, Studio Winner, 2023 Architizer Vision Awards, Photorealistic Visualization For years, renderings were treated as the final step in the process. Once the design was complete, someone would generate a few polished visuals to help sell the concept. They weren’t exactly part of the design conversation, but rather, a way to illustrate it after the fact. That’s no longer the case, however. The best rendering artists are involved early, helping shape how a project is perceived and even how it develops. This results in visualizations that don’t just represent architecture, but influence it, affecting crucial decisions in the process. Through framing, atmosphere and visual tone, renderings can set the emotional register of an entire design, meaning that rendering artists have a much bigger role to play than before. Image by Lunas Visualization, Special Mention, 2023 Architizer Vision Awards, Architectural Visualizer Of The Year Recognizing this shift in studio culture and design thinking, the Vision Awards treats rendering as its own form of architectural authorship, capable of shaping how buildings are imagined, remembered and understood. To reflect that sentiment, the program includes categories that celebrate mood, meaning and precision alike: Photorealistic Rendering – For visuals that bring spatial clarity and technical realism to life. Artistic Rendering – For painterly, stylized or interpretive representations. Architecture & Atmosphere – For renderings that evoke emotion through light, weather or tone. Technology Expanded the Medium Image by iddqd Studio, Special Mention, 2023 Architizer Vision Awards, Architectural Visualizer Of The Year Rendering used to be a time-consuming process with limited flexibility. Now, however, entire scenes can be generated, re-lit, re-textured and even redesigned in mere minutes. Want to see a project at dawn, dusk and golden hour? You can. Want to swap out a concrete façade for charred timber without starting from scratch? That’s part of the workflow. But these new capabilities are not limited to speed or polish. They open the door to new kinds of creativity where rendering becomes a tool for exploration, not just presentation. What if a building had no fixed scale? What if its context was imagined, not real? Silk & Stone by Mohammad Qasim Iqbal, Student Winner, 2023 Architizer Vision Awards, AI Assisted Visualization And then, of course, there’s AI. Whether used to generate inspiration or build fully composed environments, AI-assisted rendering is pushing authorship into uncharted territory. The results are sometimes surreal, sometimes speculative, but they speak to a medium that’s still expanding its identity. The Vision Awards recognizes these new roles of visualizations, offering categories for rendering artists that focus on experimenting with tools, tone or technique, including: AI-assisted Rendering – For images that push the boundaries of representation using generative tools. Artistic Rendering – For stylized visuals that embrace abstraction, mood, or imagination. Context Became a Key Part of the Picture Image by BINYAN Studios, Special Mention, 2023 Architizer Vision Awards, Architectural Visualizer Of The Year Architecture doesn’t exist in isolation and, increasingly, neither do the renderings that represent it. By showing how a design sits within its surroundings (whether it’s a busy street, a lakeside, or a forest) visualization becomes a way of understanding context, not just composition. In this new era of visualization, renderings show where people gather, how light travels across a building, or what it feels like to approach it through trees, traffic or rain. Movement, interaction and use-cases are highlighted, allowing viewers to grasp the idea that architecture is more than a single object, but rather, a part of a bigger picture. Image by Lunas Visualization, Special Mention, 2023 Architizer Vision Awards, Architectural Visualizer Of The Year That shift comes from a growing awareness that design is experienced, not just observed. A rendering can communicate density or calm, movement or pause, the rhythm of a city or the quiet of a field. It can reveal how a project sits in its environment or how it reshapes it. The Vision Awards includes several categories that speak directly to this expanded role of rendering, including: Architecture & Urban Life — For renderings that depict street-level energy, crowds, or civic scale. Architecture & Environment — For visuals grounded in landscape, terrain, or ecosystem. Exterior Rendering — For exteriors that communicate architectural form through environment, setting and scale. Architecture & People — For moments that highlight human presence, interaction, or use. Details Tell the Story Natura Veritas by David Scott Martin, Special Mention, 2023 Architizer Vision Awards, Photorealistic Visualization New tools have made it easier to render with nuance by highlighting texture, light and atmosphere in ways that feel specific rather than generic. With real-time engines, expanded material libraries and refined lighting controls, rendering artists are spending more time on the parts of a project that might once have gone unnoticed. Image by ELEMENT, Studio Winner, 2023 Architizer Vision Awards, Architectural Visualizer Of The Year This shift reflects changing priorities in architectural storytelling. Material choices, interior qualities and subtle transitions are becoming central to how a space is communicated. Whether it’s the grain of unfinished timber or the glow of morning light across a tiled floor, these moments give architecture its tone. The Vision Awards includes categories that reward this level of focus, recognizing renderings that carry weight through surface, rhythm and mood: Exterior Rendering — For close-up visuals that highlight the materials, textures, and design details of a building’s outer skin. Interior Rendering — For immersive representations of interior space. Architecture & Materiality — For images that showcase texture, depth and construction logic. Rendering Is Architecture’s Visual Language — and the Vision Awards are Here to Celebrate It Cloud Peak Hotel above the Rainforest Mist by FTG Studio / Zhiwei Liu, Xianfang Liu, Special Mention, 2023 Architizer Vision Awards, AI Assisted Visualization Architectural rendering is no longer a supporting act. It’s a growing creative field with its own voice, influence and momentum. As visualization continues to shape how projects are developed, discussed and shared, it’s clear that the people creating these images deserve recognition for their role in the architectural process. The Vision Awards were built to recognize exactly this. By highlighting both the artistic, technical and conceptual strength of architectural imagery, the program gives visualization the space it’s earned — alongside architecture itself. If you’re an Arch Viz artist, you can explore multiple categories that reflect the challenges, innovations and opportunities of this new era of rendering—from photorealism to abstraction, mood to material. And if your work reflects a strong point of view across multiple images, the Rendering Artist of the Year accolade was created with you in mind. Winners are featured across Architizer’s global platforms, published in print, included in the Visionary 100 and celebrated by a jury of industry leaders. Winning means visibility, credibility and long-term recognition at a global scale. So if your work helps shape how architecture is seen and understood, this is your platform to share it (and, hopefully, your time to shine!). Enter the Vision Awards Got a project that’s too bold to build? Submit your conceptual works, images and ideas for global recognition and print publication in the 2025 Vision Awards! The Main Entry deadline of June 6th is fast approach — submit your work today. The post Is Rendering Is the New Sketch? The Rise of Visualization in Architecture Today appeared first on Journal.

You can now generate photorealistic images in Microsoft Copilot, which lets you customize and edit the visuals it creates.

The pace of AI progress is showing no signs of slacking. Following ChatGPT's big image upgrade a few weeks ago, it's now Google's turn to show off new models for generating videos and pictures from text prompts: We've got Veo 3 (for video) and Imagen 4 (for pictures), announced during Google I/O 2025, and they come with some significant improvements.Starting with Veo 3, it's the next step up from the Veo 2 model that was recently pushed out to paying Gemini subscribers last month. Google says Veo 3 brings with it notable improvements in real-world physics (something AI video often struggles with) and details such as lip-syncing. In short: Your clips should look more realistic than ever.There's another crucial upgrade here, and that's sound. Previously, Veo-made clips came without any audio attached, but the AI is now smart enough to add in suitable ambient sounds, including traffic noise, wildlife sounds, and even dialog between characters.Google has provided a few example videos to show off the new capabilities, as you would expect, including Old Sailor. Of course, it's impressive that a clip like this can be produced from a text prompt, and it is up to a high standard in terms of realism—we're no longer getting the six-fingered hands that we used to with AI. Still, the usual hallmarks of artificial intelligence are evident: This is a generic sailor, on a generic sea, speaking generic dialogue about the ocean. It's a mashing together and averaging out of every video of the sea and old sailors that Veo 3 has been trained on, and may or may not match the original prompt (which Google hasn't given).Veo 3 is only available to those brave enough to pay $250 a month for Google's AI Ultra plan, but Veo 2 is also getting some upgrades for those of us paying a tenth of that for AI Pro. It's now better at control and consistency, according to Google, with improved camera movements and outpainting (expanding the view of a frame). It can also have a go at adding and removing objects from clips now.Moving on to images: We've got Imagen 4, the successor to Imagen 3. Here, we're promised "remarkable clarity in fine details like intricate fabrics, water droplets, and animal fur," plus support for higher resolutions (up to 2K) and more aspect ratios. You get top-tier results in both photorealistic and abstract styles, as per Google. There are sheep as big as tractors in Google's AI world. Credit: Google Google has also tackled one of the major problems with AI image generation, which is typography. Imagen 4 is apparently much better than the models that came before it in terms of making characters and words look cohesive and accurate, without any weird spellings or letters than dissolve into unintelligible hieroglyphics.Imagen 4 is available now to all users, inside the Gemini app. Google hasn't mentioned any usage limits, though presumably if you don't have a subscription you'll hit these limits more quickly, as is the case with Imagen 3 (there's no fixed quota for these limits, and it seems they depend on general demand on Google's AI infrastructure).The carefully curated samples Google has provided look good, without any obvious mistakes or inaccuracies—just the usual AI sheen. Imagen 4 is faster than Imagen 3 too, Google says, with more improvements on the way: A variant on the model that's 10x faster than Imagen 3 is going to be launching soon.There's one more image and video tool to talk about: Flow. It's an AI filmmaking tool from Google that pulls together its text, video, and image models to help you stitch together successive scenes that are consistent, featuring the same characters and locations. You can use Flow if you're an AI Pro or AI Ultra subscriber, with higher usage limits and better models for those on the more expensive plan.