Michael Sheldon's ubuntu touch Stuff

Over the Easter weekend I finally got around to implementing a first prototype of an idea I’ve had for a while, which aims to bring some state of the art computer vision techniques to mobile devices.

Deep Vision uses the implementation of convolutional neural networks provided by libccv to classify images. So it’ll try to figure out whatever is the principal object in an image your provide it with.

At the moment it just has a sample classification database from the ImageNet project, containing 1000 assorted items, however in the future I’d like to see specific classifiers for different tasks (e.g. a classifier trained purely on different plants, so when you’re out for a hike and you want to know what something is you can just point your phone at it and find out.)

Unlike something like Google Goggles it’s doing all the classification on the phone itself without needing to upload the image to any external services.

The video below provides a quick demo of it in action and you can also grab a click package here to play with it yourself: http://mikeasoft.com/~mike/com.mikeasoft.deepvision_0.1.3_armhf.click

Source code can be found at: https://launchpad.net/deepvision

It was just hacked together over the weekend, so it’s still a little rough in places but all the core functionality should work reasonably well :).

Tomorrow (Monday the 3rd of March) at 5pm UTC I’ll be giving a talk about QML and Box2D based game development for the Ubuntu App Developer Week, details of my talk can be found here: http://summit.ubuntu.com/appdevweek-1403/meeting/22144/game-development-with-qml-and-box2d/

In preparation for this I’ve put together a simple template for getting started with QML and Box2D development for both desktop Ubuntu and Ubuntu Touch. It’s available in two flavours:

• Precompiled version – Includes QML Box2D already compiled for amd64 and armhf (when I have more time I’ll add i386 to this as well).
• Source version – Makes it easy to compile everything yourself on whatever architecture you’re interested in.

This means that if your game is going to be purely QML based you can just grab the precompiled version, and run “make click-packages” and have packages built for both desktop and mobile use at the same time.

The template comes populated with an example application (one of the standard QML Box2D demos), which is found in the “src/” directory; so you’ll be able to see something running straight out of the box, then when you’re ready you can just replace this with your own game.

In the future I also plan to extend these templates to provide example packaging for multiple different QML + Box2D compatible mobile platforms (Sailfish, MeeGo, Blackberry, Android, etc.)

If you’re interested in seeing an example of the sort of thing you can achieve fairly easily with QML and Box2D I’ve also uploaded a video of one of my current work in progress projects:

Overview

CuteSpotify makes it possible to listen to your Spotify songs on Ubuntu Touch. It’s based on MeeSpot (a MeeGo Spotify client), which I’ve updated to make use of Qt5 and Ubuntu’s QML components. It’s still a little rough around the edges but most of the core functionality is in place now.

One particular problem to look out for is that because of the way Ubuntu Touch currently handles applications CuteSpotify has to be kept in the foreground and the phone has to be kept switched on for music to play (otherwise the application gets suspended). Approaches for handling applications that need to keep running are currently being debated, so hopefully that won’t be the case in future version of Ubuntu Touch.

As a temporary workaround, if your phone is in developer mode (achieved by connecting it up to the Ubuntu SDK) you can run “sudo service powerd stop” (default password is phablet) to stop the phone from going to sleep while you’re listening to music, then “sudo service powerd start” to resume normal power management when you’re done. (See popey’s comment for an alternative method).

Video

Installation

To install CuteSpotify on Ubuntu Touch simply search for “CuteSpotify” on your phone and it should appear.

Source

The source code for CuteSpotify can be found here: https://github.com/Elleo/cutespotify

Overview

Eyrie is a music identification program that I originally created for the Nokia N9 but which I’ve now also ported to Ubuntu Touch. It works in a similar way to commercial applications like Shazam and SoundHound, listening to music through the phone’s microphone and then generating an acoustic fingerprint using the open source EchoPrint algorithm. Currently the EchoPrint database isn’t anywhere near as large as the various commercial offerings, so it won’t recognise everything, but it has reasonable coverage.

Video

Installation

To install Eyrie on Ubuntu Touch simply search for “Eyrie” on your phone and it should appear.