Love of Computer Graphics + Machine Learning + Photography = A Spot in the iPhone Apps Hall of Fame
The display of an iPhone measures just 3.5 inches diagonally, yet as millions of iPhone fans will attest, it’s a portal into a vast world of apps, some 350,000 and counting at present. For the developers of those apps, it’s the space and place where inventiveness and savvy come together. For Hendrik Kueck, founder of Pocket Pixels of Vancouver, BC, the iPhone display represents the intersection of his passion for aesthetics and his intellectual interest in problem-solving, mapped to his desire to merge the two in fun and functional apps for consumers.
Born and raised in Germany, as a teenager Hendrik was fascinated with computers. Particularly interested in computer graphics, he played around with software that allowed him to create photorealistic renderings of imagined scenes. “It was Pixar-like animation, but of very simple things,” he says with a laugh. By the time he finished high school, though, these initial dabblings were enough to motivate a search for a university with an excellent computer graphics program, which he found in the University of Erlangen. Entering the program in 1996, Hendrik found a great group of researchers, one of whom was Wolfgang Heidrich, now a professor and Dolby Research Chair in UBC’s CS department. After completing the first part of his program, Hendrik decided to travel abroad to finish his diploma. In 2001 he arrived at UBC, began working on his thesis project in computer graphics with Dr. Heidrich, and enrolled in the UBC Master’s program in computer science. Within two years he had finished writing his German diploma thesis and was undertaking a switch in research interests.
By this point Hendrik had become interested in machine learning, a branch of computer science where algorithms are designed to detect patterns in data so as to have a predictive value for the future. As an example, he notes, a researcher can collect gene data both from people who have cancer and those who don’t, and then use an algorithm to compare the two, decide which genes are indicative of cancer, and use this data to predict who might develop cancer in the future. This research field became compelling enough that it formed the basis for his Masters’ degree, which he completed in 2004, supervised by professor Nando de Freitas.
At that point he enrolled in the Ph.D. program. He was unsure about becoming an academic, but he loved the intellectual challenges posed through research, and he was finding the CS department at UBC a fascinating place to be. “The atmosphere is extremely collegial and friendly, it is great! UBC has lots of top-notch researchers. In my two areas, in graphics and machine learning, UBC is definitely world class. And of course Vancouver is a fantastic place to be!”
His initial love of computer graphics and his growing interest in machine learning led Hendrik to work on an array of problems, and as a committed amateur photographer, he was particularly interested in problems that might help other photographers. In one project, he examined large datasets of photographs on the web to see if any rules constituting a good photo could be distilled. His hope was to take user comments on photos and see if any computer algorithm could discover some of the more subtle rules around why a particular composition was deemed good. If the algorithm was successful, perhaps a photo-cropping device or other in-camera method of assisting photographers could be invented.
The second project was oriented toward a subfield in CS called Optimal Experiment Design, which addresses the problem of choosing the parameters of a scientific experiment such that the experiment will yield as much new and useful information as possible. Hendrik worked for a time researching efficient algorithms for finding such optimal experimental designs, but he began to realize that for him, this type of research didn’t hold enough appeal because it was largely focused on the theoretical side of things. “I’m fascinated by the theory, and solving the problems is very stimulating, but I need the connection to reality as well. Working on synthetic example problems is not very satisfying for me. I feel I need an actual problem that I’m solving that has real world relevance. I’m definitely more motivated by the applications. And that was missing for me.”
A more applied research project was to come in the summers of 2006 and 2007, when Hendrik did internships at Adobe. There he worked to develop a keywords program for photographers. At the time existing programs worked by showing the photographer keywords that had been used most or most recently, but they didn’t link in any way to the photo. With Hendrik’s program, which ultimately became part of Adobe’s Photoshop Lightroom, photographers could organize and retrieve images through keywords suggested based on the photos’ metadata, such as the linked information on time, date, and place the photo was taken, as well as the keywords that the photographer had previously used to describe his or her photos. Hendrik noted that although the product didn’t employ sophisticated machine learning algorithms, he found it satisfying because it was tied to a product that many people could use and enjoy.
At this point Hendrik was feeling that his motivation with the Ph.D. wasn’t what it could be. He loved research but didn’t want to stay in academia in the long term and while he loved the challenge of theoretical problems he really wanted to create things of practical value to others.
Then in a classic “life is what happens while you’re making plans” kind of moment, the iPhone revolution hit. Although Hendrik describes himself as “not a gadget person at all,” and although he wasn’t an early adopter of the iPhone, he found himself increasingly drawn to the device. Apple had always been known for giving its customers an incredible user interface experience, and the iPhone was no exception. In fact, as Hendrik saw all of the new UI conventions, the multi-touch screen, the camera, the numerous sensors, he wondered, “How could you use this for photography? The touch screen would be amazing for photo editing. You could directly interact with the photograph. With the desktop, it’s always this indirect thing where your hand controls the mouse and you have a cursor on the screen. It might feel intuitive to us because we’re so used to it, but it’s really not. If you can actually touch the photo and edit it with your finger, that’s going to be so much more intuitive.”
The following year Apple announced the App Store for the iPhone. From a developer’s point of view, the creation of this online store, where developers could present their work and customers could buy what they wanted, was nothing short of revolutionary. “In the past both the barriers for the developers and the users were very high. You (the developer) had to create your own website, distribute the software and find a way to have credit card processing, protect yourself with a license key. You had to make sure that people find your website, that it shows up on Google, you had to advertise.”
On the user end, things were just as daunting. “The users need to find this website, trust this website, pull out their credit card, download something, get an email, copy a code, etcetera. It’s a big hassle, which is why most people don’t do it; they download a pirated version or they don’t download software at all. The App Store changed all that. Your app is going to be visible on every single device. It’s very easy to buy it and Apple takes care of the whole payment process.”
Having developed an appreciation for and interest in user interface design, he was keen to try his hand at designing intuitive and efficient UIs. Seeing that many of the walls between creative software development and access to the marketplace were being erased, Hendrik knew he should jump. He saw that the iPhone’s interface would allow for some interesting photography apps, so he began by experimenting with some “hobby projects…to create something that people use and like.”
Five months later, Hendrik had created Juxtaposer, a photomontage app that allows users to stack two photos, erase parts of the top photo and combine the photos into an often-humourous montage. Seeing the possibilities inherent in the software underlying Juxtaposer, he later created a second app called Color Splash based on the same source code. In Color Splash, a photo is converted to black and white and the user can then ‘paint’ back the original color into parts of the image, for example to highlight a specific person or object. While Juxtaposer did well enough in the Apps Store, Color Splash really took off and quickly became very popular. Before he knew it, Color Splash rose to the third spot on the list of best-selling iPhone apps. Based on user votes on the “Best App Ever” website, it won the Best Photomanipulation App of 2009, was picked by Apple as one of the Best of 2009, and ultimately became so popular that Apple chose it as one of its 50 all-time great apps, in 2010 placing it in the App Store Essentials Hall of Fame: http://itunes.apple.com/WebObjects/MZStore.woa/wa/viewRoom?fcId=402496709&mt=8&id=25204
In 2009, Hendrik created Pocket Pixels, his iPhone and iPad app development business, and in 2010 he left the Ph.D. program. Making the move from academia to running his own business has proven a “fascinating learning experience” for Hendrik. There are downsides to be sure, and he notes that “some parts I enjoy less than others, dealing with accounting being one of them.” But the positive aspects of being self-employed are numerous. For one, he has considerable flexibility in his work and can devote as few or as many hours as he wants to an array of projects of his choice. Running a one-person company has also provided exciting learning opportunities. In addition to iPhone programming, Hendrik has enjoyed learning about user interface design, graphic design, video editing, and to his surprise, even such things as marketing, branding, and pricing strategies.
One of the biggest upsides of the industry right now, what makes it so exciting to be an apps developer for the iPhone and iPad, is that the nature of the business has undergone such a radical change in just a few short years. Prior to the opening of the App Store, Hendrik would have found himself working hard on those aspects of the business that had little to do with content development and those far away from his passions: product distribution, payment processing, advertising and search engine optimization. But Apple’s arrangement with developers has changed all that. Today Hendrik and others in the field spend much of their time working on the apps themselves, networking with each other, brainstorming new product ideas, and connecting with their customers. In a kind of “democratization” of this corner of the economy, small producers play on the same field with large software development firms because the App Store has given them a venue for doing so. “It’s a fantastic time to be an entrepreneur in this space,” says Hendrik, who has developed a lot of collegiality with other developers, whom he finds very supportive and interested in sharing and growing together.
In thinking about working with the best collaborators, Hendrik’s gaze returns to UBC, where he lived in a shared house with other friends in the CS master’s program. Several of these friends actually re-grouped years later, started an independent game company and developed “Osmos,” a game that also became a big hit on the App Store and that Apple named “iPad Game of the Year” in 2010. Hendrik and his former housemates remain in close contact and often exchange tips and insights about the App Store and about running independent businesses.
Finding your passion, following the germ of an idea into another, better idea, and then another: these days Hendrik spends his time in ways that mirror his time at UBC. There, he worked with top-notch professors and continually refocused his goals and refined increasingly interesting problems for himself in an intellectually challenging setting. Hendrik counsels other graduating CS students to embrace this process of learning and growing, saying that it’s important to “figure out what you’re passionate about. And follow that. While persistence is important both for finishing a degree and for shipping a product, don’t be afraid to change things up as your passions change. Change is good!”