Brains and machines: Back to neuromorphic engineering

Screen Grab from EETimes.After some time off to focus on teaching at UCL, in the last couple of years I’ve been starting to write about neuromorphic engineering (and other topics related to intelligent machines) again. This started at the beginning of 2018 when I wrote a case study on photonics in neuromorphic systems for my book on reporting on emerging technologies. Last year I wrote a feature for Nature Electronics delayed special issue on neuromorphic computing (now scheduled to come out this summer) and this year I’ve started writing for EETimes again. I commissioned a special project on the subject for them and have also started writing a regular column on Brains and Machines.

My passion for this subject is greater than ever, and I am working on a book on this subject. More as the work develops.

 

A silicon vocal tract

Analog VLSI & Biological Systems GroupResearchers at MIT have developed the first integrated-circuit vocal tract, which could eventually make it’s way into high-end PDAs. It’s biologically inspired and combined with a bionic-ear processor in a feedback loop. This means it can not only be used for producing speech but also recognizing it: the vocal tract can help to model what the ear thinks it is hearing to verify whether it’s likely to have it right or not. Read More …

IBM cognitive computing project: so far so good…

Image from the IBM cognitive computing project.I don’t know much about this widely-publicized project to build a electronic brain in a brain-sized package as I’ve just found out about it and I haven’t yet managed to get my hands on any real technical information. However, reading the IBM press release I found out something very positive that has been omitted from most of the coverage so far. One of the researchers collaborating with IBM is Kwabena Boahen, head of Stanford’s Brains in Silicon Lab, a former student of Caltech’s Carver Mead, and a key researcher in the neuromorphic engineering community. Exactly what this means in terms of the approach that the researchers are taking is hard to say, but I’ll do my best to find out.

Stretching electronics

Stretch: EE Times cover, 22 September 2008.I’ve been interested in on-chip and chip-to-chip communications for many years, partly because optics looked likely to be an important technology in this area, and partly because dense interconnectivity is important for neural systems. However, you may be interested in an advance in interconnects for a different purpose: allowing electronics to stretch and conform. The UIUC team that did the work have used it to build a hemispherical detector, and the technology may well have an important impact on wearable sensors and actuators.

One of the impressive things about this work is how quickly it’s evolved: in the space of a couple of years they’ve gone from lines of silicon with a little give, to fully-functional very-stretchable circuits and systems. Not surprisingly, the team have already started up a company to exploit the new technology.

I know I’m not the first to write about this work, but the other stories I’d read didn’t really explain the technology or the importance to any serious extent. I hope you like the piece.

Photo: EE Times cover on 22 September 2008.

Developing systems, challenging assumptions

Mary Lou Jepsen with the  XOI was on the phone today with Mary Lou Jepsen, founding Chief Technology Officer of One Laptop Per Child and now founder of Pixel Qi, a commercial spin-off company of OLPC that will be putting their new ambient-light-viewable displays into cell phones and laptops. Mary Lou and I both started off in holography about twenty years ago, and have (miles permitting) been friends for most of that time. And she has never ceased to amaze me: both in her talent and in her fortitude. Read More …

Connecting in 3D

Irvine Sensors scheme for 3D interconnection back in 1998.I’m currently reading Ray Kurzweil’s book, The Singularity is Near, which I’ll properly review later. Among other things, the book talks about the supposed imminence of our being able to simulate the brain. I’m afraid I’m not convinced by his arguments. Don’t get me wrong: it’s not that I think it’s not going to happen. It’s just I really think he minimizes the engineering challenges that will have to be overcome to make it happen. I’ve been interested for many years in the challenge of building brain-like hardware and it’s not (to say the least) a trivial problem. Read More …

Letting the physics do the thinking

Although I’ve no doubt that digital computing will be crucial to the development of intelligent robotics, one of my interests is in the other—often neglected—technologies that will also be vital to making it happen. One of these is mechanics. The video shown here (top) is Domo, one of the latest robots being developed at the MIT Computer Science and Artificial Intelligence Laboratory. As well as incorporating many new ideas, this robot builds on one concept that was developed at the AI lab a decade ago: the idea of compliant limbs.

Read More …

On Intelligence by Jeff Hawkins

On Intelligence by Jeff HawkinsI have mixed feelings about what I consider to be ‘celebrity’ popular science books: being big in Silicon Valley and having something sensible to say about intelligent machines are two different things. In my view, however, Jeff Hawkins has paid his dues and deserves to be taken seriously. Though there is a lot wrong with On Intelligence, I believe Hawkins central theses—that much of what we call intelligence is based on the ability of our neocortex to make predictions, that the function of the neocortex is basic across sensing and actuation, and that this function can be understood in a relatively straightforward way—are all correct. If we’re smart, some of us in the machine intelligence and neuroscience communities will take up his challenge to work on filling out this new theory. Read More …

An Oxford Charade?

CharadeSaw an interesting talk yesterday. The Visual Geometry Group from Oxford University claimed to have developed a technique for automatically tagging video content (both arbitrary objects and faces) so that it was easily searchable in real time. The faces could even be labelled as particular actors or characters using online resources. The goal was to have a video version of Google where you can search by names, faces, logos, object images, anything. Although searching is not an area I particularly follow, the work seemed to have important implications for machine vision and I was amazed at how effective it was as described in the presentation. Read More …

Making the best of what you’ve got

This holographic systems pulls out individual sounds from noise using feedback. Click the image to hear this happen.<p data-wpview-marker=

This holographic systems pulls out individual sounds from noise using feedback. Click the image to hear this happen.

When I discuss analog computation with most people, their instant reaction is to explain to me that there is no such thing as analog. For instance, one of the comments on an earlier post I wrote on analog vision pointed out that using a film-based camera rather than a digital camera doesn’t give you infinite resolution, just a higher, finite resolution. In this case the the limit is grains of silver on the film rather than the number of pixels on a camera array. This is true, but it also misses the point. Read More …