Unsupervised learning of 3D structure from images

the morning paper

Unsupervised learning of 3D structure from images

Unsupervised learning of 3D structure from images Rezende et al. (Google DeepMind) NIPS,2016

Earlier this week we looked at how deep nets can learn intuitive physics given an input of objects and the relations between them. If only there was some way to look at a 2D scene (e.g., an image from a camera) and build a 3D model of the objects in it and their relationships… Today’s paper choice is a big step in that direction, learning the 3D structure of objects from 2D observations.

The 2D projection of a scene is a complex function of the attributes and positions of the camera, lights and objects that make up the scene. If endowed with 3D understanding agents can abstract away from this complexity to form stable disentangled representations, e.g., recognizing that a chair is a chair whether seen from above or from…

View original post 774 more words

Screening C# Candidates: Let’s Play 20 Questions!

Tony Sneed's Blog

Over the past year I was involved in the process of interviewing candidates for both mid and senior level developer positions. We would bring them in for a face-to-face interview, sometimes with multiple interviewers, only to find out they were unable to answer the most basic technical questions concerning C# and .NET. I’m of the persuasion that every .NET developer should understand basic concepts, such as C# language syntax, inheritance, generics, memory management, threading, etc. Without such an understanding, a developer can end up writing apps that are shot-through with problems that make the code difficult to debug and maintain.

Furthermore, I’m looking for a developer with a thirst for knowledge. The technical landscape is constantly shifting, and developers need to come up to speed quickly if they are going to be able to leverage latest enhancements to the platform. In addition, the breadth of the technical spectrum is…

View original post 1,027 more words

The Holy Trinity

Existential Type

The Christian doctrine of trinitarianism states that there is one God that is manifest in three persons, the Father, the Son, and the Holy Spirit, who together form the Holy Trinity.   The doctrine of computational trinitarianism holds that computation manifests itself in three forms: proofs of propositions, programs of a type, and mappings between structures.  These three aspects give rise to three sects of worship: Logic, which gives primacy to proofs and propositions; Languages, which gives primacy to programs and types; Categories, which gives primacy to mappings and structures.  The central dogma of computational trinitarianism holds that Logic, Languages, and Categories are but three manifestations of one divine notion of computation.  There is no preferred route to enlightenment: each aspect provides insights that comprise the experience of computation in our lives.

Computational trinitarianism entails that any concept arising in one aspect should have meaning from the perspective of the other…

View original post 926 more words

“Cows Have to be Milked”

Pittsburgh (Abolitionist) Vegan

I have come across this argument before, but not in several years. Today, two fellow students reminded me of this common argument: “Cows have to be milked.”

Searching the Internet, I came across conflicting information. Some sources claim that the cows may contract Mastitis, an inflammation of the mammary gland. I also came across several sources that state cows become agitated when not milked. Others still, claim that the teats will eventually just dry up. I have contacted the Holstein Association for more information. (Holsteins are the dominant breed of dairy cow in the United States.)

[Update 05/23/12] At the time I originally published this article, I did not have a response from the Holstein Association. Now I do:

“Hi Pierce,

While we are not a veterinary organization, which would be the best source, I have lived on a dairy farm my entire life and can hopefully help clarify it…

View original post 688 more words

A Dramatic Tour through Python’s Data Visualization Landscape (including ggplot and Altair)

Regress to Impress

Why Even Try, Man?


I recently came upon Brian Granger and Jake VanderPlas’s Altair, a promising young visualization library. Altair seems well-suited to addressing Python’s ggplot envy, and its tie-in with JavaScript’s Vega-Lite grammar means that as the latter develops new functionality (e.g., tooltips and zooming), Altair benefits — seemingly for free!

Indeed, I was so impressed by Altair that the original thesis of my post was going to be: “Yo, use Altair.”

But then I began ruminating on my own Pythonic visualization habits, and — in a painful moment of self-reflection — realized I’m all over the place: I use a hodgepodge of tools and disjointed techniques depending on the task at hand (usually whichever library I first used to accomplish that task1).

This is no good. As the old saying goes: “The unexamined plot is not worth exporting to aPNG.” 

Thus, I’m using my discovery…

View original post 5,025 more words

Video Conference Part 1: These Things Suck

Ben Garney

What I cannot create, I do not understand. – Richard Feynman

I do a lot of video chat for work. If it’s not a one on one, it’s pair programming. If it’s not pair programming, it’s a client meeting. I use a lot of Skype and Hangouts.

Sometimes they don’t work for unclear reasons. Sometimes file transfers fail. Sometimes screenshare breaks, or when it’s active you don’t get webcam, too. Or the connection lags or drops even though everything is running fast.

Every time I experience such a failure, I get really angry and think, “I could do this better!” But I never quite got angry enough… until now. I guess the weight of years of frustration finally got to me.

I wrote my own (prototype) video conferencing app. It turned out pretty well. And that’s what these posts are about.

Conventions & Caveats

We will be referencing a 640×480…

View original post 933 more words

The 100:10:1 method – the heart of my game design process

Nick Bentley Games

design-method

This is the second post of a series on practical game-design techniques. Here’s the first

In my years designing games, my methods have evolved from Games-Randomly-Emerging-from-the-Inchoate-Chaos-of-my-Brain-Area to something resembling an honest-to-goodness, write-downable process. I’ve decided to share this process here, for four reasons:

1. I’ve used it to create 3 of my 4 favorites among my own designs (Catchup, Stinker, and Cat Herders – Odd is the exception), which suggests it might have value.

2. I haven’t seen anything exactly like it.

3. Writing about it will give me ideas for improving it.

4. Pondering game design is one of the two great pleasures of my life (the other is spending time with my ladylove, who’s just sort of discombobulatingly great to be around)

Recombobulation Area …so thank heaven for this

I call it the 100:10:1 method. I’ll start by describing it, then discuss why it helps me.

The 100:10:1…

View original post 2,061 more words