Beautiful hand-drawn block diagrams from plain text
diagrams  drawing 
june 2017
[1610.02357] Xception: Deep Learning with Depthwise Separable Convolutions
We present an interpretation of Inception modules in convolutional neural networks as being an intermediate step in-between regular convolution and the depthwise separable convolution operation (a depthwise convolution followed by a pointwise convolution). In this light, a depthwise separable convolution can be understood as an Inception module with a maximally large number of towers. This observation leads us to propose a novel deep convolutional neural network architecture inspired by Inception, where Inception modules have been replaced with depthwise separable convolutions. We show that this architecture, dubbed Xception, slightly outperforms Inception V3 on the ImageNet dataset (which Inception V3 was designed for), and significantly outperforms Inception V3 on a larger image classification dataset comprising 350 million images and 17,000 classes. Since the Xception architecture has the same number of parameters as Inception V3, the performance gains are not due to increased capacity but rather to a more efficient use of model parameters.
google  deeplearning  neuralnetwork 
may 2017
Ask HN: Did your life as a parent affected your life as a developer? | Hacker News
I have a full time job as full stack developer, wife and 2 kids. I commute every day 1 hour in each direction. On a work day I woke up at 5:30 am, wash, dress and jump in a car at 6:00 am, so that I can be at work at 7:00 am. I usually go home at 3:00 pm and if there is no traffic jam I am at home at 4:00 pm. I eat something, change clothes, clock says 5 pm and I try to spend some quality time with my wife and kids. Kids go to bed at 8 pm and at 9 pm I am exhausted as hell, and I fell to bed. And this repeats every work week.

On weekends I barely have time for my side projects (one is familyokjobs.com, which I created over several weekends when wife and kids were sleeping), because there is always something to do around the house. What I'm trying to say: where do you take time for your side projects or studying new technologies?
parenting  programming  hn 
may 2017
Why Use Postgres? | Hacker News
Comment son the article about what's new in postgresql
may 2017
Enabling Continual Learning in Neural Networks | DeepMind
Computer programs that learn to perform tasks also typically forget them very quickly. We show that the learning rule can be modified so that a program can remember old tasks when learning a new one. This is an important step towards more intelligent programs that are able to learn progressively and adaptively.
ai  machinelearning  ml  neuralnetwork 
may 2017
Show HN: Sorting Two Metric Tons of Lego | Hacker News
A thread about a self-built machine that sort bulk lego parts into bins using machine learning.
machine-learning  lego  opencv 
april 2017
duraflex product,buckle,ring,tensionlock for backpack,apparel
all the plastic parts that are on backpack and messenger bar straps seem to be referenced by this company.
messengerbag  backpack  spareparts  replacementparts 
april 2017
Modern SSH server for teams managing distributed infrastructure
enterprise  security  ssh  secure 
april 2017
Syncthing replaces proprietary sync and cloud services with something open, trustworthy and decentralized. Your data is your data alone and you deserve to choose where it is stored, if it is shared with some third party and how it's transmitted over the Internet.
data  filesystem  distributed  dropbox 
march 2017
GitHub - monostable/awesome-electronics: A curated list of awesome resources for electronic engineers and hobbyists
Electronic engineering (EE) is the practice of understanding, designing and building electronic circuits. It is often differentiated from electrical engineering in that it mostly deals with low power DC electronic circuits rather than high power AC systems but there is a lot of overlap between electronic and electrical engineering.

Experimenting with and building electronic circuits is also a popular hobby and many professional resources are often equally applicable to hobbyists and vice versa.

This list is for websites, services, software, tools and more: everything that you think is awesome in the world of electronic engineering. If you have anything to add please follow the instructions in contributing.md.
awesome  circuit  diy  electronics  reference 
january 2017
OuterVision Power Supply Calculator
Power useage calculation for estimating the required wattage of a computer power supply unit (PSU)
hardware  tech  tool 
january 2017
GitHub - soumith/convnet-benchmarks: Easy benchmarking of all publicly accessible implementations of convnets
Easy benchmarking of all public open-source implementations of convnets. A summary is provided in the section below.

Machine: 6-core Intel Core i7-5930K CPU @ 3.50GHz + NVIDIA Titan X + Ubuntu 14.04 x86_64
Imagenet Winners Benchmarking

I pick some popular imagenet models, and I clock the time for a full forward + backward pass. I average my times over 10 runs. I ignored dropout and softmax layers.
ml  cuda  nvidia 
january 2017
GPU folks: we need to talk about control flow – Medium
An exploration of bugs in shader etc in the iPad / iPhone PowerVR GPU
3d  webgl  powervr  ipad  ios  iphone 
december 2016
turbo.js - Easy GPGPU
GPGPU made simple.

Offload complex parallel calculations to the GPU and get JS-accessible results
with a single call. Crunch tens of millions of numbers at once.
data  gpu  graphics  javascript 
november 2016
Oh, shit, git!
Git is hard: screwing up is easy, and figuring out how to fix your mistakes is fucking impossible. Git documentation has this chicken and egg problem where you can't search for how to get yourself out of a mess, unless you already know the name of the thing you need to know about in order to fix your problem.
cheatsheet  git  howto  reference 
september 2016
Neural Network Architectures
Deep neural networks and Deep Learning are powerful and popular algorithms. And a lot of their success lays in the careful design of the neural network architecture.

I wanted to revisit the history of neural network design in the last few years and in the context of Deep Learning.
ai  history  ml  deeplearning 
september 2016
GitHub - adereth/dactyl-keyboard: Parameterized ergonomic keyboard
The Dactyl is a parameterized, split-hand, concave, columnar, ergonomic keyboard.
keyboard  keyboards  ergodox  diy  3dprinting 
july 2016
« earlier      
3d advice ai algorithm algorithms analytics android animation api app apple apps appstore architecture art article bash bigpicture biology blog blogs book book-or-author-to-read books brain browser business c c++ canvas career charts cli clojure cloud cocoa code coding computer concurrency cool css css3 culture data database deployment design dev development distributed diy django documentation ebook economics economy editor education electronics emacs email entrepreneurship font fonts framework free fun funny game games git github google graph graphics grid gui hack hacking hardware health history hosting howto html html5 http humor icons ideas image images infographics information inspiration interesting interface internet ios ipad iphone iphonedev java javascript keyboard language latex layout learning library life lifehacks linux lisp mac machinelearning macosx management map mapping maps marketing math mathematics mobile multitouch music network networking nice-design objective-c online opensource optimization osx pdf performance philosophy photo photography physics png politics post-to-read postgresql privacy processing productivity programming psychology python rails reference research resources robot ruby scalability science scifi screen security server sf shell shopping society sociology software space sql startup statistics svg sysadmin tech technology testing text tips tool tools toread tutorial tutorials typography ui unix usa versioncontrol video video-to-watch visualisation visualization vps web web2.0 webapp webdesign webdev webkit writing

Copy this bookmark: