This new version brings a lot of speed back, thanks to better use of memory and a better organisation of tasks. There should be less crashes too, please continue to let me know about them. Thanks to these optimisations, we were able to switch from thumbnails to previews earlier than before, which I’m sure will please many people.

There may be some issues with setup: the project now uses the Intel compiler, as it generates code optimised to a greater extent. This has improved performances but also may require more files than I was distributing previously. Please let me know urgently if you have setup problems.

Go to the dowload area and download your version now!

Please, please, please, send feedback.

There is no way gpuViewer can be explained through screenshots only and I have finally come round to doing a little video on the subject.
You’ll have to put up with my french accent and less than dynamic presentation but despite all its defects, I still do hope you will enjoy this video !

Go to the dowload area and download your version now!

Big thanks to Mathias for his help with filming and editing !
Big big thanks to my brother Denis for letting me use his collection from around the world as a test bench.

Today is an important step forward in the life of gpuViewer with the addition of the “where” view.

Now, instead of having just a calendar in your main view, you also have a map. The user interface stays the same: you simply zoom using the mouse wheel…

To put your pictures on the map, nothing complicated: go to a day in the calendar, click on the earth image at the top right corner of that day, and type in the name of a city. That day is now associated with that place and it will show on the map. Associating days with places, instead of individual photos, is the easiest way to sort your photos by location: because it is quickly done, you will actually do it. And when the time comes that you want to search for a particular picture, zoomin on a place and being quickly led to the day it was taken will usually be more than enough help. That’s the whole philosophy of gpuViewer put in practise: require minimal manual indexing work and make the most of it.

To achieve this feature, we rely on two fabulous online resources: Microsoft Virtual Earth provides magnificent satellite views of our earth and geonames.org provides a database of names that you can take with you. That’s right: with gpuViewer, you can geotag your photolibrary without being connected to the internet.

When you first launch gpuViewer v0.3, it’s geonames database is empty. You should use the “File/Import from geonames” menu, which will download to your computer important data like countries and regions names and codes as well as the names and location of all cities with a population larger than 1000.

This version is quite an early one, but given the power of the feature, I wanted to share it early. I’ll release a bit later a version that let you download more places from geonames, define your own spots and manage more easily days where you’ve been to several places.

Remember: this is all work in progress, and your feedback is expected !

Go to the dowload area and download your version now!

The SDK is finally available and it looks very good! From what I’ve seen, it does give the sort of access I was looking for to do a straight port of gpuViewer. And More !

The other good news is that it seems sufficiently close to normal OSX development that it should make it possible to develop the OSX and iPhone versions in parallel, sharing a lot of the code. This is something I didn’t dare to hope… In a way, this makes having postponed the OSX version until now a good decision: this will allow me to write things in a way that takes all aspects into account.

I am very excited today because a lot of unknowns are vanishing and we now have solid ground on which to design a system encompassing everything : workstation, laptop, handheld and server.

Thank you, Apple, for giving us such an extensive access to the iPhone and iPod Touch platform !

and it’s now available in France! (as well as 37 other countries).

This is important to photographers because one of the most neglected aspects of digital photography is having a backup strategy, let alone implementing it or keeping any regularity in this process. Most people have all their photos on one hard drive (and it’s nice we can do that), and backing up is often only about copying them to another drive. That second hard drive might be as close to the first as sitting inside the same machine or on a shelf next to it… Some people backup on DVDs or to their server over a LAN, and essentially end up with their backup sitting right next to the original in a similar fashion

Ever considered fire, flooding or robbery? Backups need to be taken off-site.

The easiest way to do this is to sync your photo library to an external hard drive that gets taken to work or to a friend’s place. But even this way, things may go wrong. Files can go corrupted without your noticing it, and it’s not unheard of to go to a backup only to realise the backup itself had been bad for a long time…

That’s where online storage comes in, according to me. It is the right place where to back your best pictures up. There is no point in trying to backup hundreds of gigabytes online: even if it wasn’t going to be expensive, it would take forever to upload (your internet connection might well be uploading 10x slower than it can download).

Online storage is the right place to store your best pictures only, the ones you don’t want to lose, whatever happens and 5GB is probably enough space for most people as this will hold several hundred pictures. And simply because it is free, it’s worth getting your skydrive and saving your precious files there. For people with needs for more space, or who want a degree of automation, it’s worth also looking at Amazon S3 in general and Jungle Disk in particular.

In the middle term, gpuViewer will feature syncing features that will allow you to keep copies of your library in sync on your external drives, on your LAN and online. Hopefully, we can totally automate your backup strategy and that’s the only way to make it truly safe.

link: Welcome to the bigger, better, faster SkyDrive!

In a very surprising session of just one day, I have both managed to integrate Intel’s Performance Primitives (to deal with JPEG both fast and in a cross-platform manner) and see a 64bits build run for the first time. One thing is obvious, here on a quadcore with 8GB RAM : we waste a lot less time accessing the disk, and things are smoother. gpuViewer doesn’t make use of vast amounts of memory yet, but this will be the massive advantage of the 64 bits version: be able to deal with a gigantic image cache.

The 32bits version remains my reference platform for the foreseeable future, but it’s nice to have the option of an x64 port for people dealing with very large datasets.

With the release of v0.2, a bit of tidying up of the documentation was in order. This is now done, you can access it here: gpuViewer user guide

The important part explains how to use the new SIFT related features, you can skip directly to that if you prefer: Photo Recognition

Please send as much feedback as possible, any feedback.
and don’t hesitate to spread the word, too !

This has been a long update to come and I hope you will all realise the sort of step forward it represents.

First of all, let’s talk a little bit of the smaller updates.
- there has been quite a bit of improvement on previous/next navigation. Using the left or right arrow keys bring the previous/next picture into view in a more convenient manner by moving and scaling the display appropriately. Reviewing a set of pictures is now much more convenient. Let me know if you have more suggestions.
- the right-click menu should be working better now, but still can take some time to show up (pretty much like in the Windows Explorer itself). Note: it works on images, not on year/months/day/group icons
- I’ve started integrating Intel’s Threading Building Blocks for new features and it works amazingly well. This will be retrofitted into the rest of the app and might make things even smoother.

Now, the main new feature: SIFT integration

SIFT is a computer vision technology that enables the computer to recognize elements in an image. The general principle is that it associates a number of keypoints that it can later detect elsewhere. This has received quite a bit of attention in the particular application of automatic panorama stitching, where it has shown great precision and convenience.

In gpuViewer, my aim is to provide a means to find a picture by providing a variation. Let’s take one simple example: I keep posting pictures online in a hurry, and sometimes I just can’t remember where the original is. With this new version, it’s only a matter of copying the web version in the clipboard and asking gpuViewer to find it. Thanks to the keypoints, it can find pictures even when they have been well photoshop’ed. On my two current test databases, which contain 20000 and 15000 SIFT-indexed pictures, the search time is in the order of a few seconds (on a QuadCore Q6600/2GB/Vista PC). This is partly thanks to SSE&multicore optimisation of the matching code but mainly down to an original indexing algorithm that I’m still tuning but that seems extremely promising. Revolutionary, even, as I don’t know of other efficient ways to index SIFT data.

The process of indexing is reasonably fast, when you take into account that, for each photo, this involves decoding it, detecting up to a thousand keypoints, storing the resulting 128K on disk and keeping indexes in place. Now, when you are looking at indexing tens of thousands of images, this can mean days… Fortunately you don’t add 10K pictures everyday in your database . To make it easier to do partial tests, the database is indexed going back in time (ie the newest stuff first). If you cancel and restart later, it does recognise what’s already been done.

Let’s look a bit into the future: what is this all really for? Surely, looking up one’s own pictures is cool, and can come in handy when someone brings in a picture that he wants an enlargement of (scan it, and let gpuViewer search), but isn’t there something cooler? Well, it all depends on how well my indexing method scales. If it scales as much as I hope, then it will be possible to have central indexes (on the web, on Intranets) that will be able to let anyone find information about anyone else’s (indexed) picture. That’s right: imagine showing gpuViewer a copy of the famous surfer+shark picture (at the bottom of this post) and be directed to Kurt Jones’ site… Where you might find more pictures, licensing info, prints ordering details, or whatever the author might want (hey, it’s his site…)

Ok, this might all sound quite crazy but it isn’t and the best way for you to test is to check it out.

Go to the dowload area and download your version now!

Attention! This version is *not* SIFT enabled by default. To enable the features, you will need David Lowe’s “SIFT demo program” and copy it in the gpuViewer install directory. Please, take note of the important notice he puts forward: This demo software is provided for research purposes only. A license must be obtained from the University of British Columbia for any commercial applications. The sofware is protected under a US patent as listed below. This demo software is a research implementation, while the licensed software has been further optimized for speed and to provide a range of other capabilities. See the LICENSE file provided with the demo software. Make sure this applies to your test !

Once you have added the siftWin32.exe file to the directory, there will be a new menu in the main menu bar and new items in the context menu for photos. I’m afraid you will have to experiment by yourself until I update the user manual for gpuViewer (hopefully, tomorrow).

As always, please, keep the feedback coming!

This is a splendid computer, with many desirable features, but not the one I was expecting most… Multitouch is there, implemented in a shy way, through the trackpad, when I was hoping for the actual screen to be multitouch. This isn’t going to be as intuitive, and my initial impression is that it isn’t going to be much more than a new way to emulate the mouse wheel on a trackpad. Which isn’t that bad at all, but far from the revolution that multitouch screens will create when they do appear. I still think Apple is in the best position to make this happen, for the same reasons I mentionned previously: they have a head start with the iPhone, plus they have control over both the OS and the hardware which makes it possible for them to make it all available in one single step.

Maybe next year, then… In the meantime, I guess I should resume gpuViewer/OSX development.

I am looking very closely into the dataportability initiative, which is well described in the video below:

DataPortability - Connect, Control, Share, Remix from Smashcut Media on Vimeo.

One set of features planned for GpuViewer is the ability to not only let you upload your pictures online, but also keep track of what happens to them, then. I’ll come back to that later as this is still on the drawing board but, in the meantime, I wanted to give my nod to the initiative mentioned above, one that should make sharing things in general, and pictures in particular, easier.