Life is a dream, no?

People will say this is “peak brainrot.” But it’s interesting to think about separating the educational content from who delivers it - and then optimizing both things for each student. These videos are striking a chord, they each get millions of views!

FormSynth, by William Latham, 1989 Latham's first experimental system for art generation, FormSynth, was a system for drawing on paper, and not a computer system. The rules, sphere, cone, bulge, scoop, and so on, were typed out, and Latham applied them by hand.

@janusch_patas Oh right in 4D case you explicity state “this camera is stationary” - yeah makes sense. I’m still stuck in 3D mindset :D

@foxofdev It is both, extrinsics and intrinsics. Colmap has no information about other frames. It process a single frame for all cameras at a time.

Here's what happens when you process each frame individually in Colmap! I discovered an important lesson: processing frames in Colmap without carrying over fixed camera parameters from one frame to the next reveals unexpected variability, even with fixed, unchanging cameras. It's a surprising finding that highlights the variability of something considered accurate and a gold standard. Maybe not shocking, but definitely eye-opening.

Human.

@poetengineer__ Which OS are u running it on? Linux?

td plugin: derivative.ca/community-post… model from polycam

splatting in td

When the GeForce 256 launched there were several tech demos that showed off the capabilities of the GPU. One of them was the "Grass" demo which could render 10,000 blades of grass and allowed you to configure the weather in real time. 👍

If you are a newbie in programming, don't touch LLMs and auto complete extensions and editors trust me you won't reach your true potential by using those struggling and finding out the answer by surfing endless sites and threads is a must for true programmer character arc

Video projection with depth estimation on a point cloud matrix

【動画生成AIにも使えるカメラワーク集】 7つのカメラワークの名称と動きをまとめたショート動画。 Arc、Pan、Track、Dolly Back、POV、Low Angle、High Angleとワードとカメラの動きをセットになっていて、とてもわかり易い。 THX:@SpellOdyssey #動画生成AI #AI

【生成AIで現実世界に書き込む】 空間ペインティングと生成AIを組み合わせたアートワーク。 スマートグラスが当たり前の世界になった時に、こうした生成AIとの組み合わせは表現のハードルが一気に下がって良いですね。 #生成AI #AI Creator：@MartinNebelong

@_philschmid I also go crazy when smalltalk is directed at me

Turns out small talk is more difficult than expected for LLM Chat models. Standard SFT datasets often fail at simple greetings and go completely crazy (screenshot). This comes from missing conversation data samples in open instruction datasets like Magpie or WebInstruct. To fix the Hugging Face Team build "everyday-conversations”, a synthetic small dataset to improve small talk. 🚀 TL;DR: 🗣️ Small talk challenges even advanced chat models 🔢 2.2k multi-turn conversations generated by Llama-3.1-70B-Instruct 🎯 Covers 20 everyday topics (100 subtopics each) and 43 elementary science topics (10 subtopics each) 🗨️ 3-4 short exchanges between User and AI Assistant starting with greetings like "Hi", "Hello", "Hey!", or "Hi there" 🚀 Models now respond appropriately to casual greetings and basic identity questions 🛠️ Generated using llm-swarm tool Dataset: huggingface.co/datasets/Huggi… Big Kudos to @LoubnaBenAllal1 and the smol LLM team at @huggingface. 🤗

💿

The antithesis of 'Procreate' (very very rough tests for a new 𝐋𝐀𝐓𝐄𝐍𝐒 tool)

@DeemosTech @_akhaliq I miss the part where its open source

MeshFormer has the best quality I have ever seen on open source community! So close to Rodin!

@Teknosa Pesin fiyati da 3 tane mac fiyatina AQ

MacBook, peşin fiyatına 6 taksit imkanı ile Teknosa’da!

strange times

@majti89 @mapo1 @LinfeiPan Is it better than VGGSfM from Facebook?

🚀 Ready to take 3D reconstruction to the next level? Whether you're working on NeRF or 3DGS, our new method, GLOMAP, is here to impress! 🌟 It's faster and more accurate than COLMAP on several datasets. 🌐 Website: lpanaf.github.io/eccv24_glomap/ @mapo1, @LinfeiPan, J. Schönberger

@tarbos6 @DanielHShaffer Just another day on the job

@DanielHShaffer Indeed.