Posted by ljmacphee on August 31, 2007 under useful websites | 
I ran across this website, Computer Programming Algorithms Directory, while looking for resources for topics on this website. It’s not very large but it’s a good collection of algorithms and source code many of which will be of interest to artificial intelligence developers.
If you are looking for quick how to do somethings use Google Code Search. I find it useful for looking up specific things I know how to do but have forgotten the details.
There is also Sourcebank which has a directory of links to source code.
And for bigger projects, many of which you can join and contribute to try Source Forge
And there’s a list to your right of source code on various topics on this site. Topics that don’t have source code here usually have the algorithms spelled out for you. See specific topics for more information.
Posted by ljmacphee on August 29, 2007 under artificial intelligence in the news, computer vision |
In a radical new approach to solving identiy theft, CBL researchers are using three-dimensional information to obtain a unique biometric signature of a person’s face. With cutting-edge hardware and novel algorithms, they are designing system that turns a process practically as effortless as taking a photograph into a powerful authentication protocol.
Remembering dozens of personal identification numbers and passwords is not the solution to identity theft. Both are inconvenient to memorize and impractical to safeguard, and in essence merely tie two pieces of information together. Once the secret is compromised, the rest follows. The solution is to be able to tie private information to its owner in a way that cannot be compromised - biometric authentication
The CBL’s URxD system has the potential to move face recognition technology to the high performance gear needed for widespread application. The system determines not only the characteristics of each face, but also whether the person is wearing glasses, allowing for a practical system which offers high accuracy. So far, face recognition methods have focused on appearance - capturing, representing, and matching facial characteristics as they appear on two-dimensional images in the visible spectrum. This is quite challenging to machine recognition because such characteristics vary with orientation, age, habits (e.g., bearded appearance), and illumination. Instead, our system uses three-dimensional information, and has achieved the best published results when tried to 4,007 datasets (part of the international face recognition Grand Challenge organized by NIST). These results show strong promise in overcoming the difficult problems that have been holding back progress in this field for many years.
Biometrics - your face is your password
Cory Doctorow: The Totalitarian Urge: total information awareness and the cosmic billiards * the mp3 has a long intro, be patient the talk is worth it. This talk presents the other side of using bioinformatics as id.
While we all want to protect our identities it will change the nature of the internet and greatly reduce the freedom we enjoy online and off. Always trade offs, we just have to decide which ones we are willing to make.
Posted by ljmacphee on August 27, 2007 under artificial intelligence in the news, robotics |
Science Daily — Engineers at Purdue University are developing robots able to make “educated guesses” about what lies ahead as they traverse unfamiliar surroundings, reducing the amount of time it takes to successfully navigate those environments.The method works by using a new software algorithm that enables a robot to create partial maps as it travels through an environment for the first time. The robot refers to this partial map to predict what lies ahead. . . .
Guessing Robots Predict Their Environments, Navigate Better
More information:
Simultaneous Localization and Mapping
PDF of slide show ‘Introduction to SLAM’
Posted by ljmacphee on August 24, 2007 under artificial intelligence in the news, robotics |
WASHINGTON: Dutch researchers at the Technical University of Delft have designed a micro airplane that mimics the swift. The swift is not only the smallest soaring bird, but can go up to three years without landing.
Named RoboSwift, the micro airplane has shape-shifting, adjustable wings that allow it to manoeuvre at very high or low speeds.
With a wingspan of about one-and-a-half feet and weighing just under three ounces, the remote-controlled Roboswift is slightly larger and heavier than a common swift. But like the common swift, the robot can change its wing shape and surface area to glide or dive.
According to its designers, the robotic bird could prove useful in conducting surveys over disaster areas, or to peek in on suspicious people. . . .
More information:
Robotic Swift that flies and spies, The Times of India
Roboswift - Bio-inspired morphing-wing micro aerial vehicle
Posted by ljmacphee on August 22, 2007 under artificial intelligence in the news |
A major challenge for pioneers in artificial intelligence has been to create a living neural network on an artificial substrate. Researching the field known as nanobiotechnology, Tel Aviv University scientists have shown that it is possible to store rudimentary memories in an artificial culture of live neurons. They are apparently the first in the world to have actually stored information in a cultured neural network for an extended period.
Published in Physical Review E last month, TAU physics Professors Itay Baruchi and Eshel Ben-Jacob report the ability to record information in a man-made network of neurons. They say this is a step toward a cyborg-like amalgamation of living material and memory chips. The research may also help neurologists understand how our brains learn and store information.
This is the first time, says Ben-Jacob, that multiple rudimentary memories have been imprinted in neural networks cultured outside the brain. . .
More information:
Living Neural Networks Could Drive War Machines
TAU takes first steps in creating living cyberbrain | Jerusalem Post
Time-invariant person-specific frequency templates in human brain activity ( 2006 article Baruchi/Ben-Jacob )
Towards neuro-memory-chip: Imprinting multiple memories in cultured neural networks ( not free $25 )