Hey there, this is the sixth edition of this newsletter where I share new things I find interesting in technology and AI (paper summaries, open-source, more) straight to your inbox!
Great reads from the communityđź“–
This paper is aimed at bridging the gap between geometric deep learning and continuous models. These are formalized as the counterpart to GNNs where the input-output relationship is determined by a continuum of GNN layers, blending discrete topological structures and differential equations.
How does physics shape flight? To show how fundamental wings are, Sam derives one from scratch by differentiating through a wind tunnel simulation.
Existing techniques typically directly regress proposals in a single feed-forward step, leading to inaccurate estimation. NeuralBF is a novel instance parameterization for top-down instance segmentation on point clouds that proposes to propose a method based on iterative bilateral filtering with learned kernels.
An open high-resolution satellite imagery dataset; curates nearly 10,000 square km of unique locations; an accompanying open-source Python package is also provided to rebuild and extend the dataset.
Modern email is a patchwork of protocols and extensions. This is a long but very interesting article to understand them all from the first principles. Many thanks to Harsh Kapadia for sharing this!
Diffusion models have shown remarkable results in image synthesis, but they might not be the best way to generate images. To incorporate the inductive bias for image data, this paper proposes a novel generative process that synthesizes images in a coarse-to-fine manner.
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
Open-source from community đź‘Ź
A new library for differentiable nonlinear least squares (NLS) that is particularly useful for applications like robotics and computer vision
Betty is an automatic differentiation library for generalized meta-learning and multilevel optimization.
That’s all, hope you liked this. Stay tuned for more updates. Feel free to submit any links for the next issue.
Regards,
Rishit Dagli