Microsoft just dropped Trellis.2 — a 4B model that converts any image into a 3D asset in 3 seconds. Textured mesh under 100ms on CUDA, outputs a GLB file ready for Blender, Unity, and Unreal. Open source. Any image. 3 seconds. THE 3D ASSET PIPELINE JUST CHANGED FOREVER.

This #CVPR2026 paper from our research team is trending #1 on @HuggingFace 🤗 Meet LocateAnything: a vision-language detection model that rethinks bounding box prediction. For AI agents and robots, “seeing” is only useful if a model can pinpoint where something is fast enough to act. Trained on 138M high-quality samples, LocateAnything decodes bounding boxes in parallel instead of one coordinate at a time, improving localization accuracy while dramatically increasing throughput for visual grounding and detection. Project page: nvda.ws/4dKSohb

せきぐちあいみさん の3D Gaussian Splattingの龍に、奈良井宿の3DGS街道を飛んでもらいました！ 開発中の3DGS自動リギングで、スプラットデータを解析→ボーン自動生成→アニメーション、すべてブラウザ内で処理しています。

One of the more interesting and thought provoking research papers I've seen in a while. A system for reading and reimplementing NeRF papers, and it seems to work very well. Pretty easy to extrapolate out from here to what CVPR 2027 papers will look like. seemandhar.github.io/NERFIFY…

Our part-aware 3D generation work, OmniPart, is accepted by Siggraph Asia 2025. Code and model released! Paper: arxiv.org/abs/2507.06165 Project page: omnipart.github.io/ Code: github.com/HKU-MMLab/OmniPar… Demo: huggingface.co/spaces/omnipa…

In this iteration, I’ve added @AIatMeta’s SAM 2 to our workflow to segment people. Lisa and I had fun testing it in our lab—striking poses in sequence, with our bodies precisely isolated for a refined Stable Diffusion effect. #MR #AI #Passthrough @MetaforDevs

Mixed Reality Diffusion prototype as a tool for exploring concepts, styles, and moods by transforming real-world surroundings into alternate realities. #MixedReality #MR #AI #StableDiffusion #Quest3

Diffuman4D: 4D Consistent Human View Synthesis from Sparse-View Videos with Spatio-Temporal Diffusion Models Contributions: • We introduce Diffuman4D, a novel diffusion model that generates spatio-temporally consistent and high-resolution (1024p) human videos from sparse-view video inputs. • We propose a sliding iterative denoising mechanism that enhances both the spatial and temporal consistency of generated long-term videos while maintaining efficient inference. • We design a human pose conditioning scheme to enhance the appearance quality and motion accuracy of generated human videos. • We plan to release our processed version of the DNA-Rendering dataset, which we believe will benefit future research in this area.

展示作品で指カメラジェスチャーで撮影する機能があるんだけど、指カメラのポーズなんてこれ一択だろと思って実装組んでたら、指の向きとか位置とかで3〜4パターンくらい人によって違ってて、事前の調査不足を実感している

📝NVIDIAなどの研究チームが「Sana」をリリース 最大4096×4096の解像度の画像を数秒以内に自動生成できるAIモデル ↓建築生成結果 解像度が高いが、手摺など直線が歪んでいる。。 nvlabs.github.io/Sana/

えー、最近の大学生の固定費になります。 ChatGPT→3200円(API料金は別途) Poe→3000円 v0→3000円 Cursor→3000円 Midjourney→1600円 Runway→2400円 Dify→8800円 Taqtiq→1250円 Canva→1250円 Notta→1300円 MaxAI→3000円 Perplexity→3200円 Create→3000円 合計:34,800円/月。 AIたちの仕事量に比べて安すぎる。。 他にもオススメのAIツールあれば教えてください👇

This has to be one of the coolest things I've seen in UEFN so far 😍 UEFN has sooo much potential Great Showcase by @FNCreate X @ZHA_News

What if you could rebuild London from scratch? @EpicGames has teamed-up with @ZHA_News to create a photorealistic version of the city that anyone can explore and edit 👉 bit.ly/3XNpKno Built as an island in Fortnite, Re:Imagine London lets players adapt the urban fabric of the city and even build new developments, rewarding them for good design and use of space. #AD #architecture #construction

Gordon Ramsay in the kitchen making a mess... AI-generated video 😂 For the first few seconds, I thought it was really him, and then the AI hallucination kicked in. This is from the new Minimax model. Credit : r/u/love1008

【2枚の写真を用いた動画生成】 異なる写真を同時にインプットしてプロンプト使用して制御。物理的な制約を超えて一つの動画につなげている。 #動画生成AI #AI

Mind-reading AI recreates what you're looking at newscientist.com/article/243…

Sora is our first video generation model - it can create HD videos up to 1 min long. AGI will be able to simulate the physical world, and Sora is a key step in that direction. thrilled to have worked on this with @billpeeb at @openai for the past year openai.com/sora

Put powerful CFD analysis into the hands of the whole team. With Automate, empower everyone to contribute at an expert level. Join our closed beta programme today: forms.gle/Fj3UmaNHYuVEz4Aj8