Machine Learning World
12.3K subscribers
85 photos
17 videos
17 files
907 links
The best of Machine Learning World
@devs_world - the best materials for developers

Our fund instagram to help homeless animals: https://www.instagram.com/ukraineanimalhelp/

Contacts: @anikishaev | creotiv@gmail.com
Download Telegram
AI for AI :) This service converts math formulas from image to Latex format and saves your time. http://mathpix.com

PS: not advertisement, just really helpfull tool
​​Some simple classical CV algorithm that is very helpful during image datasets preprocessing
http://learnml.today/opencv-automatic-cropping-and-image-warping-2
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
https://arxiv.org/abs/1910.10683
Text-based Editing of Talking-head Video
https://www.ohadf.com/projects/text-based-editing/
From 30 Dec to 5 Jan we are giving a 20% discount on our Object Detection course. The best gift is knowledge. https://learnml.today/c/xyz_ny2020
This media is not supported in your browser
VIEW IN TELEGRAM
πŸŽ„Happy new year, friends!πŸŽ„
I wish you and your families all the best. May in the new year you achieve something you didn’t even dream about!
Our team is always trying to help animals, but not always our resources is enough to cover all costs :(
But if exist problem then exist solution. So we create a page on Patreon so anyone can help animals:
https://www.patreon.com/uah

Even small donations from many people can make a big difference.
People do not stop believing in bullshit. First, they were saying that Magic Leap will create new stunning AR technology, now they are saying that Samsung developed some algorithm for rendering hi-res full-body avatars with emotions and articulation.
What you see on video is just manual 3d render(best scenario)
https://www.youtube.com/watch?v=Q6f6EXX-79w
Habitat-Sim

A flexible, high-performance 3D simulator with configurable agents, multiple sensors, and generic 3D dataset handling (with built-in support for MatterPort3D, Gibson, Replica, and other datasets). When rendering a scene from the Matterport3D dataset, Habitat-Sim achieves several thousand frames per second (FPS) running single-threaded, and reaches over 10,000 FPS multi-process on a single GPU!

https://github.com/facebookresearch/habitat-sim/