Tschery retweeted
A new blog post! I've realized that the idea of applying matrices to RGBA color vectors is not as well-known as I've thought, so I've wrote a comprehensive list of operations that can be represented this way, together with some motivation for doing so:
lisyarus.github.io/blog/post…
Tschery retweeted
how software actually works for 99% of engineers: someone way smarter than you solved a really hard problem and now you build on top of their solution like adult legos and think you're a genius
Tschery retweeted
the new ai post from @sequoia is excellent. sharing best screenshots below
here's the tweet:
the opportunity lies at the application layer.
sequoiacap.com/article/gener…
Tschery retweeted
University of Oxford announces Splatter Image: Ultra-Fast Single-View 3D Reconstruction
paper page: huggingface.co/papers/2312.1…
The Splatter Image is an ultra-fast method for single- and few-view 3D reconstruction. Training is done on 1 GPU, reconstruction is done at 38 FPS and rendering at 588 FPS.
Tschery retweeted
📣"Self-Sampling for Neural Point Cloud Consolidation" ToG 2021 paper. It doesn't require training data 🙌!
TLDR; Training a neural network on the *single* input point cloud gives a sharper, and less noisy version of the input point cloud itself! 1/4
https://teapawt.net/t.co/e7ItFR4heC
Tschery retweeted
photopea (in browser, free, photoshop clone) is staggeringly good. super fast/responsive, matches photoshop 'of old', and gets lots of important ergonomic/quality details that others (cough gimp cough) dont. I feel more at home in photopea then eg affinity designer,
Tschery retweeted
First Principles of Computer Vision by Shree Nayar.
In the era of deep learning everything, understanding the fundamentals is more important than ever!
m.youtube.com/channel/UCf0WB…
Tschery retweeted
Attention is not all you need: pure attention loses rank doubly exponentially with depth
pdf: arxiv.org/pdf/2103.03404.pdf
abs: arxiv.org/abs/2103.03404
github: github.com/twistedcubic/atte…
Tschery retweeted
Neural LOD has been accepted to CVPR 2021 as an oral presentation!! 🥳
📢 Our new @NVIDIAAI paper is up on arXiv!
I'm happy to share Neural Geometric Level of Detail, which enables the first ⚡real-time rendering of high quality neural #3d SDFs with a sparse octree feature volume and a tiny MLP!
nv-tlabs.github.io/nglod/
Tschery retweeted
You want passionate people? You want people who can overcome challenges? You want creative problem solvers?
Hire non-men and pay them well. Most of us have gone through similar experiences just for a chance to do what we love.
Tschery retweeted
i hate applying for jobs because the employers are always like: do you enjoy working in a stressful environment???? do you strive for unobtainable standards of perfection???? are you good at dealing with people that make you feel extremely uncomfortable???
Tschery retweeted
DALL-E code & notebook
github: github.com/openai/DALL-E
Tschery retweeted
Multi-Knowledge Fusion for New Feature Generation in Generalized Zero-Shot Learning
pdf: arxiv.org/pdf/2102.11566.pdf
abs: arxiv.org/abs/2102.11566
Tschery retweeted
If we held a minute silence for every person who’s died of coronavirus Taiwan would be silent for 9 minutes, New Zealand for 25 minutes, Vietnam for 35, and the UK would be silient for 77 days. Don’t you dare tell me this government “done there best”
Tschery retweeted
People love to toss around the word #manifold—but what is a manifold, really?
This lecture provides a first glimpse at manifolds, using discrete meshes to side-step formal definitions that show up in many intro textbooks on differential geometry:
youtube.com/watch?v=KZjoxwUx… #DDG2021
Tschery retweeted
We are really excited to hear from Mina Konakovic Lukovic and Maria Larsson in a fabrication-themed session at 8 AM Eastern this Wednesday!
Share with your friends! (Especially, your friends in other timezones who can catch it live!)
You can watch at youtu.be/gga-yZnOpJ8

























