We get a lot of questions about what Microsoft does with the more than $9 billion we invest in R&D every year. There’s a lot of research for sure, but most of that investment goes toward development. With 850 Ph.D.-level researchers in Microsoft Research and around 40,000 developers in our product teams, that should give an indication of how we balance that $9 billion between research and the development of shipping products. I call it small r and big D.
Of course, research doesn’t just magically find its way into development. Every year we deliberately foster the connection between research and product development through Microsoft Research TechFest, three days of Microsoft Research teams showing the rest of the company their latest discoveries. Tomorrow is TechFest Public Day, when we invite a select group of academics and other guests to see some of the projects and learn about the revolutionary technologies that are coming out in the not-too-distant future.
This year a lot of the demonstrations and exhibits focus on Natural User Interfaces (NUIs) — a topic I’ve touched on several times on this blog. We believe NUIs are ushering in an era where computers are becoming more like us, and during TechFest attendees will see demonstrations of projects such as multitouch sensing on a digital pen; 3-D scanning with a regular camera; face recognition in video; and a project titled MirageBlocks that instantaneously digitizes any object into 3-D using a Kinect sensor and 3-D projector. Over the next three days, we’ll be showing you glimpses of TechFest with videos, photos and more at http://research.microsoft.com and on Next at Microsoft.
Outside TechFest, we’ve been busy undertaking technology transfer in a wide variety of ways. Some observers have questioned how much of our research is transferred into products, and wondered if we’re getting a significant return on our investment. We’re confident that we are, and although it may not always be obvious, almost every product we ship has contributions from Microsoft Research included.
Some recent examples include:
Microsoft Research and Bing Maps
The relationship between MSR and Bing Maps has always been strong. Bing Maps is trying to do things that are very edgy and push the boundaries of mapping, so there’s a natural relationship with the kinds of things that people are doing in research. The Interactive Visual Media Group (IVM) has supplied many of the recent enhancements to Bing Maps. Destination Maps is a great example where Johannes Kopf and Michael Cohen of Microsoft Research Redmond contributed to a unique feature that uses novel algorithms and cartographic principles to create the perfect party map. Another feature you may be familiar with, the Bird’s Eye View in Bing Maps, draws on the seamless stitching of images developed by Drew Steedly of Microsoft Research.
Microsoft Research and Microsoft Office
The Office Background Removal in Office 2010 relies on probabilistic reasoning algorithms from Microsoft Research to enable you to pick an object in an image and “magically” remove the rest of the image. Previously, the only way to isolate part of an image was to use a photo editing tool and learn the process of removing portions of the image. Thanks to Microsoft Research, it’s now a simple point-and-click process.
Microsoft Research and Windows Live Essentials 2011
The Photo Fuse feature in Windows Live Photo Gallery enables you to easily combine the best parts of similar photos by “fusing” them into a single perfect shot or taking unwanted people or objects out of a scene. This capability relies on sophisticated computer vision technology from Microsoft Research. Iterating over the course of several years, Photo Fuse is based on the same high-quality image matching technology found in Photosynth, Bing image search and panoramic stitching.
FUSE Labs to Bing Social
Bing has been adding more and more social features over the past year — often through collaboration with Microsoft’s FUSE Labs. FUSE stands for Future Social Experiences and is a team of researchers focused on building new social capabilities. What started as a FUSE project code-named “Twigg” resulted in what you see today on Bing Social — integrated Twitter results with any search you perform.
Microsoft Research and Xbox Kinect
The collaboration between Microsoft Research and our Xbox team was instrumental in delivering the groundbreaking Kinect experience . This collaboration has been well-documented and includes a wide variety of technologies from our labs around the world. Skeletal tracking from our Cambridge UK and Silicon Valley Labs enables Kinect to “see” you and your body movements — work that goes back as far as 2001. User identity research from MSR Asia brought face recognition, cloth color detection, and height detection to Kinect, and MSR Redmond contributed crucial audio processing technology that enables Kinect to hear you, even in a chaotic high-noise home environment.
These are just a few examples of the technology transfer and collaboration between Microsoft Research and our product teams — a process we’ve been pursuing for nearly 20 years. In the last year alone the results of our research have shipped in products such as Bing, Office, Azure, Windows, Windows Phone, Xbox, Kinect, Internet Explorer 9, Windows Live and Surface 2.0. In some cases it may be an algorithm that helps improve our search results, in other cases it may be a technology like facial recognition that is part of Kinect.
Despite these technology transfers, we’re proud that product development isn’t the primary driver of our research. Microsoft Research teams aren’t bound by product cycles and deadlines; they undertake basic research — a rare commodity in today’s commercial world. Researchers have the freedom to push their investigations where their interests lie, and to make serendipitous discoveries and connections, working both internally and with leading academics and research organizations around the world.
Microsoft Research collaborates with medical researchers at the Ragon Institute of MGH, MIT, and Harvard; the Fred Hutchinson Cancer Research Center; the University of Washington; and the University of Oxford to design an effective vaccine for HIV/AIDS, in which techniques for machine learning are used to identify the most vulnerable parts within the immune system.
In Sao Paulo, we have worked closely with Universidade de São Paulo, Fundação de Amparo à Pesquisa de São Paulo (FAPESP, the São Paulo research foundation), Instituto Nacional de Pesquisas Espaciais (INPE, the Brazilian national institute for space research) and Johns Hopkins University to develop a sensor network that can work in the challenging conditions of a rainforest and help researchers understand how different ecosystems affect the Earth’s climate and vice versa.
The balance of basic and applied research continues to be a rich source for our product teams, because the questions that researchers ask keep coming back to the fundamentals of what people want and need from technology.
And that’s crucial, because it’s what creates Microsoft’s future. We won’t lead the industry by making incremental updates and tweaks to an existing product set. To really shape the future, we have to continue to question the fundamentals, dig ever deeper, and seek to advance the state of the art. That’s what gives Microsoft the agility to respond when the world changes — or to change the world ourselves.
Posted by Steve Clayton
Editor, Next at Microsoft Blog