NIH announces big data research and education grant recipients

Vannevar Bush and an artist's rendition of his hypothetical scientific
collaboration work station, dubbed the "memex" -- memory extender

In the 1940s, Vannevar Bush envisioned scientists collaborating using networked desktop machines he dubbed "memexes." That was science fiction, but 70 years later, the Internet has opened the possibility of medical researchers creating and sharing massive clinical and health science databases. (For an early example, check out the Personal Genome Project, which is compiling individual's genome data and detailed medical records for analysis).

The National Institutes of Health has established a program called Big Data to Knowledge with the goal of advancing health and discovery through big data, and they recently announced the recipients of the first round of their big data funding:

• Centers of Excellence for Big Data Computing
• LINCS-BD2K Perturbation Data Coordination and Integration Center
• Data Discovery Index Coordination Consortium
• Training and Career Development in Biomedical Big Data
• Development of Big Data Courses and Open Educational Resources

Government research -- agency funded, as in ARPA's funding of the ARPANet, or through presidential science initiatives -- is crucial to advances in science and technology.

The importance of government research

Blumeberg columnist Mark Buchanan has written an article with the catchy title Who Created the IPhone, Apple or the Government?, in which he argues that government sponsored research has played a major role in the invention of the iPhone and much more.

I wrote a recent post about Presidential science initiatives, including several by President Obama, and have written articles on the role of government sponsored research in developing the personal computer and the Internet:

• Before the Altair: the history of personal computing, CACM 1993
• Seeding networks: the federal role, CACM 1996

I also compiled a bibliography of early contributions in:

• The precursors of personal computing, ACM SIGSMALL/PC Notes, 1988

To retrieve these articles click here and scroll down to or search for the titles.

-----

Update 7/1/2013

In a provocative blog post, Lee Vinsel asks what new technologies might come out of the NSA surveillance activity.  Will we get innovative new technologies or are they just using established techniques of big data mining and server farms? Also -- if there are new technologies, will they belong to the US people or to contractors like Booz Hamilton, which hired Edward Snowden?

This leads me to wonder whether the increasing reliance of government at all levels on contractors will diminish socially shared innovation.

Talking heads may or may not improve presentations, but they do add to conversations

MOOCs and other online classes give us an opportunity to learn what works in presentations. For example, I've proposed studies of the effect of playback speed or including a talking head in a presentation on comprehension, retention and engagement.

I'm not sure how these studies would turn out (though I have my suspicions), but I am sure that talking heads are important in coversation.

My certainty is based on a single anecdotal experience (of no scientific value). Some friends and I meet for a weekly podcast bull-session on technology (Yet another tech show). We do the podcast using a Google Hangout.

As you see here, video thumbnails of all the participants are shown under a larger head-shot of the person who is speaking. (Switching is automatic). But, check the thumbnails and you see that the second from the left is blank. That is because one of us had a slow Internet connection that night and could not stream video. We could hear what he said, but not see his face.

If you had asked me before we did that podcast whether not seeing him while he spoke would matter, I would have said "no" -- figuring that the audio conveyed all the information in a technical discussion.

If you ask me now, my answer is "yes, for sure." It had nothing to do with information or content. We understood what our invisible colleague was saying and he participated fully, but it felt weird.

Talking to disembodied people on the phone feels natural because there is no point of reference. But, in a mixed conversation, where you can see some, but not all of the participants, the difference is palpable. I don't know if it is because of subtle clues as to who gets the floor next, some primitive fear or something else, but it feels more natural to talk with someone you can see.

Two researchable questions: do talking heads and moving hands improve learning?

I would like to see studies testing the hypotheses that seeing a talking head or a hand writing on the screen improves student comprehension and retention.

I've developed a collection of modules for my digital literacy course. Each module contains an annotated PowerPoint presentation and other material. As I find time, I record videos of the PowerPoint presentations, using the annotation as a script.

To conserve bandwidth and storage space, I only use still images of the PowerPoint slides in my videos. There is no video of me speaking. I reasoned that my students knew what I looked like, and viewers who were not in my class could see what I looked like in a photo.

But, I notice that Google (left), Coursera (right) and others include video of the speaker's face during their presentations.

Is there evidence that talking heads improve comprehension and retention?

We see a similar contrast in the screencast videos produced by Udacity (left) and the Khan Academy (right}. Udacity videos are screencasts with superimposed video showing the teacher's hand and pen while drawing and writing on the screen (with noticeable parallax). Khan produces screencasts in which only a cursor is shown.

Does seeing the hand and pen improve comprehension and retention?

(If these research questions sound interesting, see a proposal for a related study of the effect on comprehension and retention of varying playback speed and bullet point images).

Software refinement -- from research to products -- an image processing example

We have discussed the birth of image processing with Ivan Sutherland's Sketchpad program in the 1960s. As we see in this video, Sketchpad could draw, rotate, resize, copy and combine images.

Since Sutherland's research, there has been tremendous improvement in hardware technology and innumerable engineering refinements to image processing software.

Software refinements require research and the development of prototypes before they become product features. Consider, for example, the PatchMatch algorithm developed in collaboration between Princeton University and Adobe Systems. Here we see a paper describing the algorithm and some demonstration videos. The paper was presented and the videos shown at an academic engineering conference.

Successful research like the PatchMatch algorithm is subsequently incorporated into commercial products. Here you see a demonstration of tools that use PatchMatch for image retargeting, completion and reshuffling integrated into Adobe Photoshop. The tools are being demonstrated at a trade show, and it is probably safe to say they will be in the next version of Photoshop.

This is typical of engineering refinements to software. A prototype is developed in a research lab and it later becomes a product.

Researchers typically work on things that are too demanding for the technology of the day. Some of the Princeton videos are shown at 5 times actual speed because it takes time to compute the new images. By the time these features are released in Photoshop, hardware will have improved and they will execute rapidly.

Note that Sutherland's work was a pure university effort, but PatchMatch was a joint effort between a university and a company. Many significant advances have come from industrial research labs. Can you find some interesting prototypes and projects at Microsoft Research?

Computer science research leads to commercial products

Computer scientists at universities and the research labs of companies like Microsoft, Google or IBM work on new ways of doing things that may one day be used in commercial projects. In addition to theoretical studies, they often build hardware or software prototypes that test and demonstrate their ideas. (The slogan "demo or die" is common in such labs).

They also publish papers describing their work in technical journals and conference proceedings. One can get a glimpse of the future by reading those papers and attending conferences. For example, the 2008 conference of the Association for Computing Machinery Special Interest Group on Computer Graphics (SIGGRAPH for short) was just held in Los Angeles. Projects presented at SIGGRAPH foreshadow tomorrow's image and video editing programs. For a first look at features that might one day be in Photoshop or Windows Movie Maker, check these projects:

• Three-D photo viewer: combine many photos of the same object (like the statue of liberty or Notre Dame Cathedral) into a panoramic scene over which the user can pan and zoom.
• "Unwrap mosaics" for video editing: modify an object like a person's face in one still frame, and have the change propagated automatically throughout the video.
• Using photographs to enhance videos of a static scene: combine a video of a scene with a few still photographs to enhance the video by improving resolution, dynamic range, or touching up or removing objects.
• Face swapping: automatically replace faces in photographs.

You will find demonstration videos and technical papers at each of these sites.

The case of Pixar Animation illustrates the research-commerce relationship. In 1986 Pixar, then a brand new company, showed its first film, a short called Luxo Jr., at the SIGGRAPH conference. Watch Luxo Jr. and you will note things like the accuracy of moving light, shadows and reflections and the sinusoidal movement of the lamp cord when Luxo jumps. Achieving that sort of realism was state of the art research in 1986, therefore relevant for a scientific meeting. (The creation of a cute character was not science, but story telling art). The interplay between research and commerce continues -- Pixar employees presented a technical paper and a course at SIGGRAPH 2008.

Entrepreneurs and venture capitalists watch developments at research conferences like SIGGRAPH for ideas -- what are some other computer science conferences?