Pinned Tweet
Interested in how 3D vision can help the development of high-resolution weather/climate models?
In our NeurIPS 2025 spotlight paper, we show how ground-based cameras can be used to estimate physically-meaningful 3D cloud properties at a high spatiotemporal resolution.
🧵 1/7
Jacob Lin retweeted
Can MLLMs learn to reason in space by forming a 3D relational scene graph, just like humans build mental maps of what they see? 👀🤔
Meet SpatialThinker – an MLLM trained with online RL + dense spatial rewards to build and reason over scene graphs.
🧵👇[1/8]
Cloud4D is trained using physically-realistic clouds that are generated using physics-based large-eddy simulations (LES) and rendered using Blender.
Both our real-world dataset and our synthetic dataset are now available (see link in the project page below).
6/7
1
4
75
For more details and results, check out our project page and paper!
Project Page: cloud4d.jacob-lin.com/
Paper: arxiv.org/pdf/2511.19431v2
w/ @CloudsClimate @ronnieclark__
Supported by @ARIA_research
7/7
4
63
Jacob Lin retweeted
Introducing the X account for the Perceptual Intelligence and Extended Reality Lab (PIXL) at @CompSciOxford!
We’ll share discussions, interesting talks and research projects from our group.
Check our work at: pixl.cs.ox.ac.uk
📸From our spring outing @HawkConservancy
Jacob Lin retweeted
Motion blur typically breaks SLAM/SfM algorithms - but what if blur was actually the key to super-robust motion estimation?
In our new work, Image as an IMU, @ronnieclark__ and I demonstrate exactly how a single motion-blurred image can be used to our advantage.
🧵1/9
Jacob Lin retweeted
Ever wondered how 3D vision can help understand the weather and climate? Check our new work on reconstructing 3D cloud fields. We test our model using data from @Raspberry_Pi cameras observing a 10km2 area of the sky.
Project: cloud-field.github.io
w/ @jacob_lin2 @EdGrys






