Paying in, paying out

James wordlessly outlines his woes

James wordlessly outlines his woes

One of the critical, game-breaking elements of this project is getting a payment API integrated into our code that works the way we want. Stripe, PayPal, Braintree, and a handful of the other usual suspects are in consideration right now, each with their own merits, and probably none with exactly the perfect combination of features that we'd want.

Some of these limitations are those we ourselves bring to the table: we're a UK based company, and this limits many of the things we're able to do with a system like Stripe. It's frustrating, but there's no getting around it, and particularly disappointing given that Stripe has so many other features we'd want to take advantage of.

We're jugging a few options here, hope to have an answer by end of the day's coding. PayPal, despite some juggling about getting accounts are cards set up, may well prove the most viable option here, provided that we can solve the issue of verifying images as fast as credit card details.

Something new

This is something of a departure.

Passage has been impressively successful, not only in terms of popularity and market uptake, but also as a proving ground for the core Peekabu technology. Validating the performance of user-created images as recognisable tokens was always one of the key user aims of the Passage project, and the lessons learned in building the apps have been phenomenally instructive. So much so, that we're trying something new, something we've been talking about for a long time.

We're going to do a one-week build. One week of focussing our team on one, single project that uses the tools and codebase we've built up in a new way. The project will combine computer vision, mobile payments, and user-created media in a brand new way. 

It's not going to be easy. It will demand a lot more of our skills than we've put in to day. It may not work at all. But we're not giving up without a fight.

Today we're announcing Project Cheq. It's a working title. It's a work in progress. We'll write up the whole thing over the course of the week, the good and the bad. No matter what, it'll be fun.

Stay tuned. We can't wait to show you.

Peekabu 'Passage' debuts at TechCrunch Disrupt London 2014

After 20 long hours in the basement of Old Billingsgate in London, Peekabu debuted 'Passage,' the technology platform for visual authentication and password management. After a rousing response, Peekabu is on track to develop a beta of the application within the coming weeks.

Stay tuned for more details, and read the article here: 

Blurred frames

A few weeks ago, Google released a discreet camera app for its Android devices, in an effort to maintain a consistent experience across the range of devices. One of the new features added was Lens Blur, a unique approach to solving the problem of shallow depth of field on mobile cameras. I highly recommend reading their blog post about how they managed it.

The key issue revolved around the problem of a single camera with minimal lenses trying to capture depth information about a scene. The new HTC One (M8) manages this by having two lenses, one which is entirely devoted to getting depth information. But of course, for the vast majority of consumer handsets, this isn't an option.

To solve this, the function instructs the user to raise the device slowly after taking the initial shot. This motion tracks which objects and features stay in a similar position in the camera frame (distant) and which move most (near). It can then construct a depth map of the scene based on this data. Very clever.


It is not a replacement for proper lenses on an SLR, but it does produce impressive results in the right conditions. As with many single camera applications, washed-out or too-dark pixels won't yield as much data as those clearly lit, but the resulting images are impressive.

What's more impressive, from our vantage, is that this feature was rolled out to everyone with Android, for free, and gives them a very cool new function that would otherwise be a key camera feature of any new handset, all without changing one bit of hardware, all using the camera to its fullest potential. Naturally, based on what we do at Peekabu, we highly approve.