Hacker News
X / Twitter
I wrote up how I built the shitty robot so you can too. This was a fun project that will keep on giving.
Thanks to all the open weights folks out there, without whom this would not have been possible.
https://t.co/egPxlCECQg
ついにできた...
これからは、C++一行も書かずにオーディオプラグインつくれる！！
web技術でVSTのUIも音声処理も書ける🙌
わざとパツパツする音でシンセ作って、DAWでループ再生している間に、DSPのソースコード(JavaScript )更新すると、ビルドもプラグインリロードもなしで、その場ですぐ音が差し代わる...革命すぎる🔥🔥🔥🔥🔥
minWM
A Full-Stack Open-Source Framework for Real-Time Interactive Video World Models https://t.co/hxUocDWGJj
🚀 Excited to share REST3D: REconstructing physically STable and visually consistent 3D scenes from a casual single image🤳.
With REST3D, you can naturally interact with stable virtual objects through hand-based VR interactions👐.
🔗 Project page: https://t.co/1CVuGIjAVM https://t.co/59FN0f7jJm
From single images, recent 3D models can generate high-quality 3D objects, but not 3D scenes.
We introduce Extend3D, a training-free pipeline to generate a large-scale 3D scene from a single image.
Project: https://t.co/gKZ2W8fmIX
#CVPR #CVPR2026 #3DGeneration #3DScene #AI https://t.co/Eggpqh2RV3
Feed-forward 3D reconstruction methods typically predict pointmaps in camera-centric frames. But why should a camera's arbitrary orientation define the coordinate system?
We introduce G3T, a transformer that predicts pointmaps in gravity-aligned frames. Regardless of input image orientation, our method always produces upright pointmaps (see demo).
We leverage this uprightness to create G3T-Long, a submap-based reconstruction method that improves robustness on long-sequence 3D reconstruction (more on that below).
Interactive demos, code, and model weights are available on our project page.
recommended viewing. dax is a smart cookie. https://t.co/vZ8ENsZt7m
In-context learning suggests that a model has learned versatile representations. What if we use in-context learning itself as a training task for visual representations?
📣 Introducing 𝗟𝗜𝗟𝗔: 𝗟𝗶𝗻𝗲𝗮𝗿 𝗜𝗻-𝗖𝗼𝗻𝘁𝗲𝘅𝘁 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 ✨ @CVPR 2026 Oral ✨
𝗟𝗜𝗟𝗔 trains on videos without manual annotation.
Key idea: An optimal linear mapping that predicts dense cues (e.g. depth, flow), estimated on one video frame, should also predict the corresponding cues of another frame from the same video.
This yields compelling results on dense vision tasks: video object segmentation, (zero-shot) semantic segmentation and surface normal estimation.
Paper, code, models and demo: https://t.co/Xn2SgskKQ8
Joint work with @ma_sundermeyer, Hidenobu Matsuki, David Joseph Tan and @fedassa (and special thanks to David and Federico for hosting my research visit at Google).
#cvpr2026 @Google @MunichCenterML @tumcvg @TU_Muenchen
Robots can now reconstruct 3D scenes in real time from a single RGB camera.
[📍 Projects page + paper]
No depth sensor. No retraining. 30 FPS.
Researchers at the Imperial College London introduced KV-Tracker, a training-free method that makes heavy models like π³ and Depth Anything 3 fast enough for real-time tracking.
The idea is simple.
These models use global self-attention, which is powerful but computationally expensive.
KV-Tracker caches the key and value pairs from selected keyframes and reuses them for new frames.
That cache becomes an implicit scene representation.
Result:
• Up to 30 FPS
• 10 to 15x speedup
• Accurate 6-DoF tracking on benchmarks like TUM RGB-D and 7-Scenes
• Works with monocular RGB only
It also supports object-level tracking with masks and allows saving the KV-cache for later reuse.
For robotics, this reduces hardware constraints and moves real-time 3D perception closer to practical deployment.
Credit to Marwan Taher (@marwan_ptr) at Imperial’s Dyson Robotics Lab and many others who contributed to this!
📍 Save projects page + paper for later:
https://t.co/IxgTl19zZ7
Video: https://t.co/wVxLhd6VpP
——-
if it matters in AI or Robotics you'll read it here first: https://t.co/9Nm01QUcw3
Excited to share our new work R³: 3D Reconstruction via Relative Regression.
Only 372M params (~⅓ of recent 1B-class baselines), trained on 6×48G GPUs, but competitive on streaming reconstruction. Runs at 30+ FPS.
Project: https://t.co/tw6oIjF5L9
Paper: https://t.co/aASJeWtxpO https://t.co/APVT5QhSgV
"RecGen: 3D Multi-Object Scene Reconstruction from Sparse Observations"
TL;DR: generative 3D scene reconstruction framework that recovers geometry, texture, and pose from sparse RGB-D observations while remaining robust to heavy occlusions and object symmetries https://t.co/hS4YDsx60Y
Flash-KMeans was only the beginning.
Today, from the Flash-KMeans team, we are releasing FlashLib — a GPU library for fast, predictable, agent-ready classical ML operators.
Up to 26× on KMeans, 19× on KNN, 40× on HDBSCAN, 208× on TruncatedSVD, 47× on PCA, 147× on exact t-SNE, and 49× on MultinomialNB over state-of-the-art (cuML).
Blog: https://t.co/P31SGl0cyT
Code: https://t.co/9nkO2hmeOl
Next video is on training tiny (<1B) models for preference tuning. Plus how to generate preference datasets with local models.
Covers reward models, RLHF, DPO, ORPO with Unsloth and TRL. Releasing sometime this week! https://t.co/iFuBj5oaIT
@ShriramKMurthi I got a tool for you. https://t.co/W1F8OEhRaM
We just released code and model! Go check it out!
Code: https://t.co/r3GvNLYgn0
Model: https://t.co/r2q8TS2zcD https://t.co/QFW08NzuwW
Really great usecase for autoresearch I've found is writing a basic nav algo (loop closure, relocalization) establishing ground truth for scoring (apriltags in this case) then autoresearching until score is perfect https://t.co/y4j3yA5Zxm
✨#CVPR2026 Oral ✨
A tale of a failed experiment: what if you fine-tune DINOv2 on sparse keypoints, beat every benchmark, only to discover it performs worse than the original frozen model on novel keypoints?
🚀MARCO closes this gap: a unified model for generalisable correspondences
https://t.co/vE62YiTVfd