⚡️ 2D → 3D Magic

It’s Friday. Turning a flat image into a textured 3D model in a matter of seconds might sound like science fiction, but Adobe researchers are making it a reality. Let’s dive in.

Today’s Highlights:

• OpenAI's quest for unbiased AI

• GitLab expands its AI lineup

• Snap Introduces ChatGPT in AR Lenses

DEEP DIVE

Adobe Researchers Turn Flat Images Into 3D Models in a Snap

A scientific leap by Adobe Research and the Australian National University has turned the tables in digital design. Their new AI model deftly transforms a 2D image into a high-quality 3D model in five seconds flat.

The paper outlines how LRM (Large Reconstruction Model) leverages massive datasets from Objaverse and MVImgNet to create neural radiance fields (NeRF) from a single image. The LRM doesn’t just generate any 3D model; it captures intricate textures with uncanny precision. Think wood grains on a chest or fabric weave—all this from a single snapshot!

What sets this tech apart is its scalability and versatility. Batch-trained on a collection of over 1 million 3D shapes and video data, its 3D reconstructions can adapt to various inputs like photos and AI-generated imagery.

Potential to reshape creative industries:

• For game developers and animators, this could mean the faster production of complex 3D environments and characters.

• Industrial designers might see accelerated prototyping, effortlessly turning 2D drafts into interactive 3D models.

• In AR/VR, imagine transforming snapshots into immersive worlds, all in real time.

While it's not perfect—some textures may go blurry, especially in occluded areas—the LRM firmly pushes the boundaries of what's possible. Its ability to generate 3D models from casual "in-the-wild" photos could democratize the creative process, making advanced modeling accessible to anyone with a smartphone.

The researchers are optimistic about the future of data-driven, large-scale 3D constructions and hope their work inspires exploration in 3D reconstructions from arbitrary images captured in natural settings.

Keep Reading Here.

PUNCHLINES

DDoS Dodgeball: Altman trolls Musk as OpenAI rolls out GPTs to all subscribers despite DDoS attack.

Pixels to Profit: Stable Diffusion creator Stability AI has reportedly raised $50 million in financing led by Intel.

Chatting Up Code: GitLab expands its AI lineup with Duo Chat, allowing devs to interact with the bot to access existing Duo features.

Cruising LLMs: OpenAI-backed startup Ghost claims LLMs will overcome self-driving setbacks.

TLDR

OpenAI's quest for unbiased AI: Acknowledging the flaws in current AI training data, OpenAI launches Data Partnerships to collaborate with organizations for cleaner, unbiased data sets. The initiative seeks diverse data reflecting all aspects of society, with partners like the Icelandic Government onboard.

Snap Introduces ChatGPT in AR Lenses: At Lens Fest 2023, Snap unveils a partnership with OpenAI to integrate ChatGPT into AR lenses, offering innovative conversational and learning experiences. Lens Studio 5.0 Beta speeds up project loading by 18x and encourages teamwork with Git integration.

Breakthrough AI invents rapid iceberg mapping: AI developed by Leeds University scientists maps monumental icebergs from space in milliseconds, outperforming humans by 10,000-fold. The U-net neural network model accurately captures iceberg outlines, aiding climate change surveillance.

Amazon's 'Olympus' aims to top ChatGPT: Amazon is set to challenge OpenAI with its own AI, Olympus, tailor-made for corporate clientele. With 2 trillion parameters, it will work alongside Amazon's retail, AWS, and Alexa services.

Picsart unveils Ignite for innovative content creation: Picsart launches Ignite, a suite of 20+ AI tools for businesses and creators, offering advanced image editing, object removal, personalized avatars, and more—aiming to simplify and enrich the content production experience.

TRENDING TOOLS

📜 Detangle: Clarify legal documents for better understanding and cost efficiency

🔍 Context by Fleet: Access high-quality embeddings for a vast library of Python tools

📊 Github Issues: Streamline project planning and tracking for development teams

📱 UI Auditor: Get AI-powered feedback on your app's user interface design

⚙️ Attio Automations: Enhance workflow with AI-driven, customizable automation

That’s all for today—if you have any questions or something interesting to share, please reply to this email. We’d love to hear from you!

P.S. If you want to sign up for the Supercharged newsletter or share it with a friend, you can find us here.