AIBITS, S.A. :: Artificial Intelligence

The Boss Is Robotic, and Rolling Up Behind You

http://www.aaai.org/aitopics/html/current.html (AAAI) — Sat, 04 Sep 2010 00:00:00 +0100

Mobile robots have been used for years by the military and law enforcement, but with falling costs, the next frontiers are the office, the hospital and the home. SACRAMENTO — Dr. Alan Shatzel’s pager beeped at 9 on a Saturday morning. Dr. Shatzel, a neurologist, hustled not to the emergency room where the patient lay — 260 miles away, in Bakersfield — but to a darkened room at a hospital here. He guided the roughly five-foot-tall machine, which has a large monitor as its “head,” into the patient’s room in Bakersfield. “Computers are beginning to grow wheels and roll around in the environment,” said Jeanne Dietsch, a veteran roboticist and co-founder of MobileRobots Inc ., a robot maker in Amherst, N.H., and a division of Adept Technologies.

Continue reading The Boss Is Robotic, and Rolling Up Behind You

Dropping Jaws and Stopping Conversations in a Test-Drive

http://www.aaai.org/aitopics/html/current.html (AAAI) — Sat, 04 Sep 2010 00:00:00 +0100

For the past three weeks I have been baby-sitting a robot. It looks like an extra-large lollipop on wheels and using it is a little like owning a puppy. For the past three weeks I have been baby-sitting a robot. It is not your traditional looking robot — if there is such a thing yet — but looks like an extra-large lollipop on wheels. People ask, “Can it do any neat tricks?” The robot I adopted, the Texai, is made by Willow Garage, a technology company based in California, and is designed for telepresence. I wheeled past, said hello over the robot’s speakers and watched through the camera as her head moved slowly with my passing, the spoon still suspended in midair.

Continue reading Dropping Jaws and Stopping Conversations in a Test-Drive

PrePrint: Kernel Optimization in Discriminant Analysis

http://www.computer.org/tpami (IEEE transactions) — Fri, 03 Sep 2010 19:44:25 +0100

Kernel mapping is one of the most used approaches to intrinsically derive nonlinear classifiers. The idea is to use a kernel function which maps the original nonlinearly separable problem to a space of intrinsically larger dimensionality where the classes are linearly separable. A major problem in the design of kernel methods is to find the kernel parameters that make the problem linear in the mapped representation. This paper derives the first criterion that specifically aims to find a kernel representation where the Bayes classifier becomes linear. We illustrate how this result can be successfully applied in several kernel discriminant analysis algorithms. Experimental results using a large number of databases and classifiers demonstrate the utility of the proposed approach. The paper also shows (theoretically and experimentally) that a kernel version of Subclass Discriminant Analysis yields the highest recognition rates.

Kernel - Operating Systems - Linux - Function - Ubuntu

Continue reading PrePrint: Kernel Optimization in Discriminant Analysis

PrePrint: Multiview Stereo and Silhouette Consistency Via Convex Functionals over Convex Domains

http://www.computer.org/tpami (IEEE transactions) — Fri, 03 Sep 2010 19:44:23 +0100

We propose a convex formulation for silhouette and stereo fusion in 3D reconstruction from multiple images. The key idea is to show that the reconstruction problem can be cast as one of minimizing a convex functional where the exact silhouette consistency is imposed as convex constraints that restrict the domain of feasible functions. As a consequence, we can retain the original stereo-weighted surface area as a cost functional without heuristic modifications of this energy by balloon terms or other strategies, yet still obtain meaningful (nonempty) reconstructions which are guaranteed to be silhouette-consistent. We prove that the proposed convex relaxation approach provides solutions that lie within a bound of the optimal solution. Compared to existing alternatives, the proposed method does not depend on initialization and leads to a simpler and more robust numerical scheme for imposing silhouette consistency obtained by projection onto convex sets. We show that this projection can be solved exactly using an efficient algorithm. We propose a parallel implementation of the resulting convex optimization problem on a graphics card. We are able to compute highly accurate and silhouette-consistent reconstructions for challenging real-world data sets. In particular, experimental results demonstrate that the proposed silhouette constraints help to preserve fine-scale details of the reconstructed shape.

Optimization - Algorithm - Function - Programming - Languages

Continue reading PrePrint: Multiview Stereo and Silhouette Consistency Via Convex Functionals over Convex Domains

'NHL2K11' shoots, but sometimes misses

http://www.aaai.org/aitopics/html/current.html (AAAI) — Fri, 03 Sep 2010 00:00:00 +0100

(CNN)-- NHL2K11 (2KSports/Visual Concepts) tries to bring realism to sports on the Wii, but falls short because of control issues and a spotty artificial intelligence. When developing their hockey game for the upcoming season, 2KSports decided to put all their eggs in one basket by only making a game for the Nintendo Wii and bypassing the Sony PS3 and Microsoft Xbox 360 consoles. By utilizing the Wii Remote controller and the Wii Motion Plus attachment, 2KSports hoped to introduce a more true-to-life feel in the game play. One-timer shots connect with power, but can sometimes be swing-and-a-miss attempts as well -- not what you'd expect from NHL style players.

Continue reading 'NHL2K11' shoots, but sometimes misses

Human unconscious is transferred to virtual characters

Thu, 02 Sep 2010 04:00:00 +0100

Virtual characters can behave according to actions carried out unconsciously by humans. Researchers have created a system which measures human physiological parameters, such as respiration or heart rate, and introduces them into computer designed characters in real time.

Continue reading Human unconscious is transferred to virtual characters

CFP: MIWAI'10 - 2nd CALL FOR PAPERS + STUDENT TRAVEL SCHOLARSHIPS

(Richard Booth) — Thu, 02 Sep 2010 01:50:04 +0100

------------------------------ ------------------------------ ---------------
Apologies for Multiple Copies Due to Cross Posting
------------------------------ ------------------------------ ---------------
============================== ==============
MIWAI'10 - SECOND CALL FOR PAPERS
============================== ==============

Continue reading CFP: MIWAI'10 - 2nd CALL FOR PAPERS + STUDENT TRAVEL SCHOLARSHIPS

PrePrint: Approximately Global Optimization for Robust Alignment of Generalized Shapes

http://www.computer.org/tpami (IEEE transactions) — Wed, 01 Sep 2010 22:44:22 +0100

In this paper, we introduce a novel method to solve shape alignment problems. We use gray-scale "images" to represent source shapes, and propose a novel two-component Gaussian Mixture (GM) distance map representation for target shapes. This asymmetric representation is a flexible image-based representation, which is able to represent different kinds of shape data including continuous contours, unstructured sparse point sets, edge maps, and even gray-scale gradient maps. Using this representation, a new energy function based on a novel two-component Gaussian Mixture distance model is proposed. The new energy function was empirically evaluated to be a more robust shape dissimilarity metric that can be computed efficiently. Such high efficiency is essential for global optimization methods. We adopt and modify one of them, the Particle Swarm Optimization (PSO), to effectively estimate the global optimum of the new energy function. Different from the original PSO, several new strategies were employed to make the optimization more robust and prevent it from converging prematurely. The overall performance of the proposed framework as well as the properties of each algorithmic component were evaluated and compared with those of some state-of-the-art methods. Extensive experiments and comparison performed on generalized 2D and 3D shape data demonstrate the robustness and effectiveness of the method.

Optimization - Particle Swarm Optimization - Algorithm - Function - Artificial Life

Continue reading PrePrint: Approximately Global Optimization for Robust Alignment of Generalized Shapes

PrePrint: Single Image Haze Removal Using Dark Channel Prior

http://www.computer.org/tpami (IEEE transactions) — Wed, 01 Sep 2010 22:44:22 +0100

In this paper, we propose a simple but effective image prior - dark channel prior to remove haze from a single input image. The dark channel prior is a kind of statistics of outdoor haze-free images. It is based on a key observation - most local patches in outdoor haze-free images contain some pixels whose intensity is very low in at least one color channel. Using this prior with the haze imaging model, we can directly estimate the thickness of the haze and recover a high quality haze-free image. Results on a variety of hazy images demonstrate the power of the proposed prior. Moreover, a high quality depth map can also be obtained as a by-product of haze removal.

Model - Graphics - Hazing - High school - St. Paul High School

Continue reading PrePrint: Single Image Haze Removal Using Dark Channel Prior

PrePrint: Direct Method for Video Based Navigation Using a Digital Terrain Map

http://www.computer.org/tpami (IEEE transactions) — Wed, 01 Sep 2010 22:44:21 +0100

A novel vision-based navigation algorithm is proposed. The gray levels of two images, together with a Digital Terrain Map (DTM) are directly utilized to define constraints on the navigation parameters. The feasibility of the algorithm is examined both under a simulated environment and using real flight data.

Navigation - Algorithm - Data - Geomatics - Earth Sciences

Continue reading PrePrint: Direct Method for Video Based Navigation Using a Digital Terrain Map

PrePrint: The Cues in "Dependent Multiple Cue Integration for Robust Tracking" are ...

http://www.computer.org/tpami (IEEE transactions) — Wed, 01 Sep 2010 22:44:21 +0100

A methodology for integrating multiple cues for tracking was proposed in several papers. These papers claim that, unlike other methodologies, conditional independence of the cues is not assumed. This brief communication 1. refutes this claim and 2. points out other major problems in the methodology.

Methodology - Business - Research - Psychology - Health care

Continue reading PrePrint: The Cues in "Dependent Multiple Cue Integration for Robust Tracking" are Independent

PrePrint: Reliability Fusion of Time-of-Flight Depth and Stereo for High Quality Depth Maps

http://www.computer.org/tpami (IEEE transactions) — Wed, 01 Sep 2010 22:44:20 +0100

Time-of-flight range sensors have error characteristics which are complementary to passive stereo. They provide real time depth estimates in conditions where passive stereo does not work well, such as on white walls. In contrast, these sensors are noisy and often perform poorly on the textured scenes for which stereo excels. We explore their complementary characteristics and introduce a method for combining the results from both methods that achieve better depth accuracy than either alone. In our fusion framework, the depth probability distribution functions on depth from each of these sensor modalities are computed and then effectively merged. Robust and adaptive fusion is built on a pixel-wise reliability weighting function calculated for each method. In addition, since time-of-flight devices have primarily been used as individual sensors, they are typically poorly calibrated. We introduce a method that substantially improves upon the manufacturer's calibration. We demonstrate that our proposed techniques lead to improved accuracy and robustness on an extensive set of experimental results.

Sensor - Fusion power - Stereophonic sound - Physics - Nuclear

Continue reading PrePrint: Reliability Fusion of Time-of-Flight Depth and Stereo for High Quality Depth Maps

Robotsdeveloped in Plymouth to befriend sick children

http://www.aaai.org/aitopics/html/current.html (AAAI) — Wed, 01 Sep 2010 00:00:00 +0100

Robots which have memories and can interact with children are being developed by the University of Plymouth as part of a European project. The Now robots are being created as pets to interact and befriend children with long-term illnesses. The University is one of eight institutions across Europe working on the 8.3m Euro project over four years. Dr Tony Belpaeme from the University of Plymouth said: "The original prototypes were going to be rather simple, in the sense that they were going to play simple games with the children.

Continue reading Robotsdeveloped in Plymouth to befriend sick children

What the papers say

http://www.aaai.org/aitopics/html/current.html (AAAI) — Tue, 31 Aug 2010 00:00:00 +0100

Journalist Fionola Meredith takes a look at what is making the headlines in Tuesday's newspapers. Several papers lead with the accident at Cranfield beach in which a six-year-old boy was seriously injured. The Irish News,The News LetterandThe Mirrorall carry reports on how the child is fighting for his life after being hit by a speedboat while he was being towed on a wake-board near a busy beach. The Irish Timesreports on an opportunity for would-be nuns.

Continue reading What the papers say

University of Plymouthrobotfootballers take on world

http://www.aaai.org/aitopics/html/current.html (AAAI) — Tue, 31 Aug 2010 00:00:00 +0100

A football team made up of robots developed by the University of Plymouth is aiming to beat the best from around the world. Tony Belpaeme, a lecturer at the university's Centre for Robotics and Neural Systems, is sending a team of 10 robots to the competition in Bangalore, along with about six students. Mr Belpaeme said: "Anything that computers can do well, such as calculations, is hard for humans. "It is going to be fun and exciting," said Mr Belpaeme.

Continue reading University of Plymouthrobotfootballers take on world

Skitterbot creeps us out at 1 foot per second

http://www.aaai.org/aitopics/html/current.html (AAAI) — Tue, 31 Aug 2010 00:00:00 +0100

Billed as the world's fastest micro-robot, Skitterbot is perfect for scaring friends who loathe insects, especially robotic ones. USB controllable. Skitterbot: Plug 'n crawl If you love freaking out your friends, especially a girlfriend who hates bugs, here's a toy you might dig--Skitterbot is a six-legged robotic insect that can burn rubber at the rate of one foot per second. Maker Desk Pets claimsSkitterbotis the fastest commercially available legged micro-robot. Skitterbots follow a horde of micro-robot toys including theHexbug Crab, which can respond to its environment through sensors. Still, I'd love to see a whole floor crawling with Skitterbots.

Continue reading Skitterbot creeps us out at 1 foot per second

CFP: A Special Issue on Emerging Synergies of Artificial Intelligence and Software Engineering, ...

(alexliu) — Mon, 30 Aug 2010 02:50:42 +0100

========= Call for Papers =========
A Special Issue on Emerging Synergies of Artificial Intelligence and
Software Engineering
International Journal of Software Engineering and Knowledge
Engineering
[link]
=== Theme ===
As software engineering is requested to answer dynamic, automated,

Continue reading CFP: A Special Issue on Emerging Synergies of Artificial Intelligence and Software Engineering, IJSEKE

EmotionML: Will computers tap into your feelings?

http://www.aaai.org/aitopics/html/current.html (AAAI) — Mon, 30 Aug 2010 00:00:00 +0100

"Affective computing" standards and technology could help computers grasp human emotion. An intuitive user interface sounds nice, but what about information overshare? For all those who believe the computing industry is populated by people who are out of touch with the world of emotion, it's time to think again. TheWorld Wide Web Consortium (W3C), which standardizes many Web technologies, is working on formalizing emotional states in a way that computers can handle. EmotionML combines the rigor of computer programming with the squishiness of human emotion. " But theMultimodal Interaction Working Groupthat's overseeing creation of the technology really does want to marry the two worlds.

Continue reading EmotionML: Will computers tap into your feelings?

In Search of Time

http://www.aaai.org/aitopics/html/current.html (AAAI) — Mon, 30 Aug 2010 00:00:00 +0100

Ever since Einstein, physicists have been telling us that time – this steady tick-tock of the universe – is much weirder than we think. It doesn’t flow in a single, linear direction, or beat like a steady metronome. (Eagleman is also a best-selling novelist –Sumis a brilliant riff on the possibility of an afterlife. “We asked everyone how scary it was, on a scale from 1 to 10,” he reports, “and everyone said 10.” And all of the subjects reported a slow-motion effect while falling: they consistently over-estimated the time it took to fall.

Continue reading In Search of Time

Video: Yale's Grab Lab Demonstrates an Unmanned Helicopter With a Grabbing Hand

http://www.aaai.org/aitopics/html/current.html (AAAI) — Mon, 30 Aug 2010 00:00:00 +0100

Researchers at Yale's Grab Lab aren't about to let the nuances of rotary-wing flight restrict what unmanned aerial vehicles can do. A team there has developed a hand-like modular grasping and manipulation platform that can be fitted to the bellies of UAVs to provide them with extra functionality without overly taxing their flight capabilities. Aerial vehicles that could pick up and carry objects for their human handlers are generally of the vertical take-off and landing (VTOL), which means they are generally small rotary-wing aircraft (helicopters). To circumvent these issues, Grab Lab engineers created their manipulator with a single motor that controls all four "fingers" simultaneously.

Continue reading Video: Yale's Grab Lab Demonstrates an Unmanned Helicopter With a Grabbing Hand

Sweet Talking Your Computer

http://www.aaai.org/aitopics/html/current.html (AAAI) — Sat, 28 Aug 2010 00:00:00 +0100

Why people treat devices like humans; saying nice things to a machine to protect its "feelings." When BMW introduced one of the most sophisticated navigation and telematics systems into its 5 Series car in Germany a decade ago, it represented the pinnacle of German engineering excellence, with great advances in accuracy and functionality. While this might seem like a story of men's weird attachment to cars or gender stereotyping run amok, a growing body of research suggests that there is something much deeper at work: People respond to computers and other technologies using the same social rules and expectations that they use when interacting with other people. These weren't overly sensitive people: They were graduate students in computer science and electrical engineering, all of whom insisted that they would never be polite to a computer. One of the most reviled software designs of all time was Clippy, the animated paper clip in Microsoft Office.

Continue reading Sweet Talking Your Computer

Kokoro shows off its latest android Actroid F

http://www.aaai.org/aitopics/html/current.html (AAAI) — Sat, 28 Aug 2010 00:00:00 +0100

Tokyo entertainment firm Kokoro shows off its latest fembot, Actroid F, in a PR video. The lifelike android is slated to go on sale to work as a receptionist or hospital worker. Geminoid F, the uncannily lifelike fembot we sawin April, is back in a new PR vid from Kokoro, a Tokyo-based entertainment company that collaborates with Osaka University'sHiroshi Ishiguroin the creation of androids bothfeminineandcreepy. In the new video, the robot calls itself "Actroid F," as it has joined the ranks of otherActroidrobots produced by Kokoro. While Actroid F can move its eyes, mouth, head, and back, it can also act as a telepresence robot. ATRIntelligent Robotics and Communication Laboratories, backed by the government, companies, and academia, also collaborated in Actroid F's development, one of many robot projects Japan has funded as it tries to develop next-generation machines to meet social needs.

Continue reading Kokoro shows off its latest android Actroid F

Sticky Robot Climbs Up Glass

http://www.aaai.org/aitopics/html/current.html (AAAI) — Sat, 28 Aug 2010 00:00:00 +0100

Mimicking a gecko's super sticky toes is no easy feat, but these scientists are pretty close. Their latest generation of a wall-climber, the Stickybot III, was inspired by the gecko lizard, which has phenomenal capabilities for climbing any vertical surface, including glass. The animal's toe contains hundreds of ridges called lamellae, and on each of those ridges are millions of tiny hairs called setae. Looking closer, the ends of the setae have even smaller strands called spatulae.

Continue reading Sticky Robot Climbs Up Glass

Use Microsoft Surface to Control a Swarm of Robots With Your Fingertips

http://www.aaai.org/aitopics/html/current.html (AAAI) — Fri, 27 Aug 2010 00:00:00 +0100

A sharp-lookingtabletop touchscreencan be used to command robots and combine data from various sources, potentially improving military planning, disaster response and search-and-rescue operations. Mark Micire, a graduate student at the University of Massachusetts-Lowell, proposes using Surface, Microsoft's interactive tabletop, to unite various types of data, robots and other smart technologies around a common goal. Indefending his graduate thesisearlier this week, Micire showed off a demo of his swarm-control interface, which you can watch below. To test his system, Micire steered a four-wheeled vehicle through a plywood maze.Control This Robot With a Touchscreen: Mark Micire/UMass Lowell Robotics Lab The system can integrate a variety of data sets, like city maps, building blueprints and more.

Continue reading Use Microsoft Surface to Control a Swarm of Robots With Your Fingertips

UsingRobotsTo Clean Up Oil Spills

http://www.aaai.org/aitopics/html/current.html (AAAI) — Fri, 27 Aug 2010 00:00:00 +0100

An MIT lab proposes a quicker, more efficient cleanup system. We may not have figured out how to stop oil spills from happening, but the folks at SENSEable City Laboratory at MIT are one step closer to a solution to cleaning up the aftermath. Their newest project, a swarm of robots called SeaSwarm, proposes to skim the ocean surface and remove oil. The first prototype was successfully tested in Boston's Charles River in August, and it responded well to the water's changing surface. Another perk of SeaSwarm is that using the flock of robots eliminates the need for human and equipment coordination in a large-scale cleanup effort.

Continue reading UsingRobotsTo Clean Up Oil Spills

EEG Monitoring Headband Could Track and Catalog Your Emotional Response to Movies

http://www.aaai.org/aitopics/html/current.html (AAAI) — Fri, 27 Aug 2010 00:00:00 +0100

Plenty of human-gadget interfaces can let you control a robot or a computer with your mind. An EEG-based affective computing system allows you tocommunicate your emotions, adding a new layer to human-computer interactions. In the video below, Robert Oschler of Android Review demonstrates EmoRate, a software program that catalogs his emotions. The computer tracks four primary emotions -- happiness, sadness, anger and fear -- and catalogs when they occur.

Continue reading EEG Monitoring Headband Could Track and Catalog Your Emotional Response to Movies

This Week in the Future, August 23-27, 2010

http://www.aaai.org/aitopics/html/current.html (AAAI) — Fri, 27 Aug 2010 00:00:00 +0100

Our beverage-powered alien robot overlords discipline us with pain rays. Welcome to the future. This week in the future onPopSci: Leave a comment (any comment) to put your name in the pile; we'll randomly choose and announce our winner right here. And, if you just can't wait that long, you can buy the shirt for yourselfhere.

Continue reading This Week in the Future, August 23-27, 2010

Surgical robot could be used for long-distance regional anesthesia

Thu, 26 Aug 2010 04:00:00 +0100

An existing surgical robot could be used to perform complex regional anesthesia procedures -- in theory, allowing expert anesthesiologists to perform robot-assisted procedures from remote locations, according to a new study.

Continue reading Surgical robot could be used for long-distance regional anesthesia

Is music going back to its roots?

http://www.aaai.org/aitopics/html/current.html (AAAI) — Thu, 26 Aug 2010 00:00:00 +0100

Technology has paved the way for almost anyone to hit the right notes are some people pining for a more authentic sound? Its 21st century equivalent is arguably the smartphone and instead of "pressing down a special key" to "play a little melody" you can launch an app to compose an orchestral concerto. This is the last of our five-part series exploring the intersection between technology and culture There is no doubt that technology has transformed the world of music makers, but has it taken away the craft? "All technology has a place, and if it's used for artistry, if the sound is beautiful, I'm all for it," he said.

Continue reading Is music going back to its roots?

Researchers Create Ultra-Sensitive Robotic Nose Using Frog Eggs as an Olfactory Sensor

http://www.aaai.org/aitopics/html/current.html (AAAI) — Thu, 26 Aug 2010 00:00:00 +0100

Researchers at the University of Tokyo are using frog eggs to enhance what might seem like an unlikely element of robotics:olfactory sensing. By injecting the eggs with the DNA from various insects known for expressing keen senses of smell, the team was able to create a robotic nose that can detect molecules at levels as low as a few parts per billion. By injecting the eggs with DNA from fruit flies, silk moths, and diamond back moths, the bioengineers were able to coax the immature eggs into producing the olfactory sensors of those insects, which are known for their ability to detect specific gases and pheromones. The eggs were then fitted between two electrodes, creating a sensor system that measured the current generated when certain molecules came in contact with the detector.

Continue reading Researchers Create Ultra-Sensitive Robotic Nose Using Frog Eggs as an Olfactory Sensor

MIT unveils swimming, oil-cleaning robots

http://www.aaai.org/aitopics/html/current.html (AAAI) — Thu, 26 Aug 2010 00:00:00 +0100

(CNN)-- Here's a new way of looking at oil spill clean-up: Forget the big ships, massive work crews and hefty price tags. Instead, just deploy an army of autonomous, oil-scrubbing robots. The $20,000 robots will be unveiled officially to the public on Saturday at an event in Venice, Italy, and will be ready to deal with oil spills in about a year, said Assaf Biderman, who oversaw MIT's research team on the project. "You can imagine it like a carpet rolling on the surface of the water," said Biderman, who also is associate director of theSENSEable City Lab.

Continue reading MIT unveils swimming, oil-cleaning robots

Kaspar the friendly robot helps autistic kids

http://www.aaai.org/aitopics/html/current.html (AAAI) — Wed, 25 Aug 2010 00:00:00 +0100

A child-size humanoid bot with a face inspired by comics is being used to teach interaction skills to kids who may lack an instinctive understanding of social cues. A toy tambourine in Kaspar's lap adds to its perceived playfulness. A playful child-size humanoid robot with a face inspired by comics and Japanese Noh theater is being used to help teach autistic children social skills. They built the robot for $2,500 using a child-shaped mannequin for the body's base, off-the-shelf parts, and silicone-rubberRoboSkinwith embedded tactile sensors that detect different kinds of touch. Kaspar is part of theAurora Project, which is aimed at investigating the possible use of robotic systems as therapeutic or educational tools for autistic children.

Continue reading Kaspar the friendly robot helps autistic kids

SIAI Newsletter June-July 2010

(Michael Anissimov) — Tue, 24 Aug 2010 00:07:26 +0100

Where to Find Us

The Singularity Institute has been extending its web and outreach presence these past two months. As always, you can find us at the SIAI homepage and at the SIAI blog.

Hip readers may want to follow us via Twitter (@singinst) or Facebook Causes. Recently, we have set up a Crowdrise crowdsourcing and crowdfunding account for volunteers and fellows and a YouTube channel to host our collection of videos – with a few future surprises.

New Leadership in the Visiting Fellows Program

The Institute’s team of Visiting Fellows is under new management. Former Visiting Fellow Jasen Murray replaced Anna Salamon as Program Manager of the Visiting Fellows Program, with Louie Helm serving as his Deputy Program Manager.

Anna Salamon spent the past year managing the expanding Visiting Fellows Program. Freed from logistics and management, she aims to spend more time on her duties as a Research Fellow, writing research papers and engaging Visiting Fellows and other visiting academics in relevant research. She is also helping Visiting Fellows develop their rationality and their understanding of the existential risks landscape through individual tutoring and group workshops.

Jasen Murray founded the New York Less Wrong meet-up group and spent two years researching gene therapy and personalized medicine before joining the Visiting Fellows Program. His focus as program manager is on identifying individuals’ unique strengths, and on creating the institutions and skills that can let Fellows coordinate well, build effective teams, and get their strengths in contact with academic research, IA projects, and other action toward a positive Singularity.

Louie Helm is the founder of the Seventeen or Bust distributed computing project and an internet entrepreneur. As Deputy Program Manager, he manages programming projects, outreach projects, grant writing, and team-building for the Visiting Fellows. Louie also manages the SIAI volunteer program and is always looking for more online volunteers to assist SIAI in its various projects: reach him at louie.helm@singinst.org.

FQXi Grant Proposals Submitted

SIAI submitted three grant proposals for a total of $183,000 to the Foundational Questions Institute (FQXi). FQXi allocates grants to research projects that it believes will tackle “questions at the foundations of physics and cosmology, particularly new frontiers and innovative ideas integral to a deep understanding of reality but unlikely to be supported by conventional funding sources.”

These grants, if awarded, would allow SIAI to fund research in several topics of longstanding interest. Among these topics are a developed timeless decision theory (TDT) that can be viewed as a proper alternative to causal decision theory (CDT), an exploration of fundamental priors and the mechanics of the Universal Distribution with Absolute Self-Sampling Assumption (UDASSA), and the implications of the simulation hypothesis on cosmology and fundamental physics.

New Program Manager Jasen Murray has promised that the institution’s grant-writing project will be the first of many whole-team projects within the Visiting Fellows program. Discussed possibilities for the future are an electroencephalograph-building (EEG) project, a book planning project, and various IA experiments.

Research: ECAP Papers and Presentations

SIAI continues to contribute to the field of AI safety and to exchange dialogue, via published papers and presentations, with other scholars and students in the field. These past two months have been particularly busy with academic work.

Carl Shulman (currently in Australia) gave two academic presentations on what the evolution of intelligence implies about the feasibility of AI, with special attention devoted towards anthropic arguments. The first was an hour-long talk at the Australian National University; the second was a conference talk at the Association of Australian Philosophers.

The European Conference on Computing and Philosophy (ECAP) accepted Anna Salamon’s newest conference paper, written in conjunction with Research Fellow Steve Rayhawk and Visiting Fellow Janos Kramar, “How Intelligible is Intelligence? Implications for AI Development Trajectories.” In this paper, they propose that one factor that can usefully constrain models of AI development and takeoff trajectories is the “intelligibility of intelligence”—the extent to which efficient algorithms for general intelligence follow from simple general principles, as in physics, rather than being necessarily a conglomerate of special case solutions.

Visiting fellow Kaj Sotala also had his conference paper, “From Mostly Harmless to Civilization-Threatening: Pathways to Dangerous Artificial Intelligences,” accepted by ECAP. Kaj’s paper discusses the threat level of three possible pathways for AI development: through advanced hardware capabilities, through the capability of software self-improvement, and through multiple copies of a digital AI. He will be editing his completed paper the following month in his native Finland.

Three additional SIAI-affiliated papers have been accepted by ECAP: Anders Sandberg and Carl Shulman’s “Implications of a Software-Limited Singularity”; Steven Kaas, Peter Salamon, and Steve Rayhawk’s “Economic Implications of Software Minds”; and Joshua Fox and Carl Shulman’s “Superintelligence Favors Cooperation, but not Benevolence.” Steven Kaas is an alumnus of the summer 2009 Visiting Fellows program, Carl Shulman continues to work on SIAI research and projects in Australia, and Joshua Fox is a long-time supporter of SIAI and of efforts to create a positive Singularity.

Michael Anissimov’s presentation proposal “Public Scholarship for Global Catastrophic Risks” was accepted by the Society for Risk Analysis conference, to be held in Salt Lake City in December.

The Visiting Fellows Program: Bigger, Faster, Better!

The SIAI Visiting Fellows Program has grown over the past two months. Since the beginning of June, seven new Visiting Fellows have entered through the doors of the program: Peter de Blanc, Dennis Fan, Shaneal Manek, Frank Adamek, Abraham Wolk, Tim Czech, and Stanislas Sochacki. We expect to welcome eight more fellows over the course of the summer.

Currently, the new fellows are filming a Less Wrong lecture series, assisting in SIAI outreach efforts, and spending time learning from the Visiting Fellows curriculum. For more information on each individual fellow and his/her interests, visit the Visiting Fellows page.

Next time…

Look forward to more exciting developments from SIAI in the following newsletter… a Less Wrong YouTube lecture series is in the works by Visiting Fellows Peter de Blanc, Dennis Fan, and Frank Adamek… and even more Visiting Fellows will be trickling in.

Continue reading SIAI Newsletter June-July 2010

Babies Take the Wheel of DrivingRobots

http://www.aaai.org/aitopics/html/current.html (AAAI) — Tue, 24 Aug 2010 00:00:00 +0100

Giving disabled babies the power of movement could improves their development. But it's the control mechanism on this particular baby robo-chair that's the real genius: by sticking the baby into a seat strapped onto a Wii Fit Balance Board perched atop a Pioneer 3 robot platform, researchers from Ithaca College created a robot that travels in whatever direction a baby leans. From the paper : The balance board has responsive pressure sensors in each of its four corners and built in Bluetooth capabilities. C++ library (ti) to access balance board data within our robotics software. Our software compares the values of the four pressure sensors to determine which, if any, direction the infant is leaning.

Continue reading Babies Take the Wheel of DrivingRobots

Willow Garage's PR2 Robot Achieves Another Laundry Landmark, This Time Cleverly Pairing Your Socks

http://www.aaai.org/aitopics/html/current.html (AAAI) — Tue, 24 Aug 2010 00:00:00 +0100

The guys over at Willow Garage have already proved their PR2 to be a worthy weekend companion, teaching it toplay billiards,fetch beer, and even to clean up after a party. A team of UC Berkeley researchers more interested in domestic applications for robotics has also shown that PR2 can be a handy household companion during the slower parts of the week, namelylaundry day. Now, they've shown that if you give PR2 a sock it can employ its keen ability for repetitive hand motions to that other regularly recurring chore:pairing socks! Presented with two socks, PR2 must first determine if each one is correctly oriented or if one or both are inside out.

Continue reading Willow Garage's PR2 Robot Achieves Another Laundry Landmark, This Time Cleverly Pairing Your Socks

Five must-have video games for fall

http://www.aaai.org/aitopics/html/current.html (AAAI) — Tue, 24 Aug 2010 00:00:00 +0100

Editor's note:Scott Steinberg is the head of technology and video game consulting firm TechSavvy Global as well as the founder of GameExec magazine and Game Industry TV. (CNN)-- From downloadable titles like haunting puzzle-action hybrid "Limbo" for Xbox 360 to the long-awaited release of PC sci-fi strategy juggernaut "StarCraft II," summer's been good to gaming fans. These games should soon prove there's little reason to make the upgrade yet, with so much life left in your PC and favorite console. Here are five upcoming titles that top my personal wish list: "Halo: Reach"(September 14; Microsoft, Xbox 360): This sequel builds upon the same futuristic run-'n'-gun thrills players know and love by adding special powers (invulnerability, temporary flight, etc.

Continue reading Five must-have video games for fall

Personalized Life Extension Conference, October 9-10, SFO Marriott

(Michael Anissimov) — Mon, 23 Aug 2010 23:57:39 +0100

Chris Peterson, a friend and advisor of SIAI, is holding a personalized life extension conference on October 9-10 at the San Francisco Airport Marriott. Readers of SIAI blog can get $100 off on registration by using the registration code “SIAI”. Here’s the short summary of the purpose of the conference:

Advances are being made daily on what each of us can do NOW to slow the aging process to a minimum, and to delay or prevent the diseases of aging.

Life extension news comes out faster than any one of us can evaluate it on our own. Let’s get together and determine how to take personal action.

Speakers include Esther Dyson, Peter Thiel, Terry Grossman, Bruce Ames, Gregory Benford, Patri Friedman, Sonia Arrison, Greg Fahy, and many more.

Continue reading Personalized Life Extension Conference, October 9-10, SFO Marriott

Singularity Summit Australia- Sept. 7, 11, 12, Melbourne

(Tom McCabe) — Mon, 23 Aug 2010 04:44:12 +0100

The first ever Singularity Summit Australia will be held on September 7th, 11th, and 12th at Casey Plaza at RMIT, Melbourne. Here’s the news release:

Singularity Summit 2010 in Melbourne Victoria to explore emerging technological trends
Speakers include AI Architect Hugo deGaris, Futurist Peter Ellyard

Melbourne Victoria 2010 — Will it be one day become possible to boost human intelligence using brain implants, or create an artificial intelligence smarter than Einstein? In a 1993 paper presented to NASA, science fiction author and mathematician Vernor Vinge called such a hypothetical event a “Singularity”, saying “From the human point of view this change will be a throwing away of all the previous rules, perhaps in the blink of an eye”. Vinge pointed out that intelligence enhancement could lead to “closing the loop” between intelligence and technology, creating a positive feedback effect.

This September 7th, 10-12th, hundreds of AI researchers, robotics experts, philosophers, entrepreneurs, scientists, and interested laypeople will converge in Melbourne to address the Singularity and related issues at the only conference on the topic in Australia, the Singularity Summit. Experts in fields including artificial intelligence, bio-ethics, tissue regeneration, medical ethics, computational neurobiology, cryonics, futurism, and more will share their latest research, brightest ideas, and explore their implications for the future of humanity.

“This is the first Singularity Summit outside the US, and will be a great event,” said Adam Ford, coordinator of the Singularity Institute Australia, which is hosting the event.

Russell Blackford, editor of “50 Voices of Disbelief” who is a contributor to IEET, will explore the ethical and practical implications of key emerging technological trends. Futurist Peter Ellyard will discuss opportunities that innovation and technology can provide and his book ‘Designing 2050′.

Interested readers can watch videos from past summits and register at http://summit.singinst.org.au.

Continue reading Singularity Summit Australia- Sept. 7, 11, 12, Melbourne

X Factor admits tweaking vocals

http://www.aaai.org/aitopics/html/current.html (AAAI) — Mon, 23 Aug 2010 00:00:00 +0100

X Factor contestants Gamu Nhengu, G&S and Shirleena Johnson audition X Factor bosses have admitted using vocal enhancement processes on the ITV talent show. Fans had claimed after the opening show that "autotuning" had been used to improve some contestants' voices. An X Factor spokesman said post-production work was necessary on the show due to the number of microphones used during filming. Somebody took the time to go, "I'm going to leave this bit dry, and then I'm going to fix that bit".

Continue reading X Factor admits tweaking vocals

PrePrint: Canonical Correlation Analysis for Multi-label Classification: A Least Squares ...

http://www.computer.org/tpami (IEEE transactions) — Fri, 20 Aug 2010 17:44:34 +0100

Canonical Correlation Analysis (CCA) is a well-known technique for finding the correlations between two sets of multi-dimensional variables. It projects both sets of variables onto a lower-dimensional space in which they are maximally correlated. It is well-known that CCA can be formulated as a least squares problem in the binary-class case. However, the extension to the more general setting remains unclear. In this paper, we show that under a mild condition CCA in the multi-label case can be formulated as a least squares problem. Based on this equivalence relationship, efficient algorithms for solving least squares problems can be applied to scale CCA to very large data sets. In addition, we propose several CCA extensions including the sparse CCA formulation based on the 1-norm regularization. We further extend the least squares formulation to partial least squares. In addition, we show that the CCA projection for one set of variables is independent of the regularization on the other set of multi-dimensional variables, providing new insights on the effect of regularization on CCA. We have conducted experiments using benchmark data sets. Experiments on multi-label data sets confirm the established equivalence relationships. Results also demonstrate the effectiveness and efficiency of the proposed CCA extensions.

Canonical Correlation Analysis - CCA - Dimension - Data - Databases

Continue reading PrePrint: Canonical Correlation Analysis for Multi-label Classification: A Least Squares Formulation, Extensions and Analysis

PrePrint: Equidistant Fish-Eye Calibration and Rectification by Vanishing Point Extraction

http://www.computer.org/tpami (IEEE transactions) — Fri, 20 Aug 2010 17:44:34 +0100

In this paper we describe a method to photogrammetrically estimate the intrinsic and extrinsic parameters of fish-eye cameras using the properties of equidistance perspective, particularly vanishing point estimation, with the aim of providing a rectified image. The estimated intrinsic parameters are the optical center and the fish-eye lensing parameter, and the extrinsic parameters are the rotations about the world axes relative to the checkerboard calibration diagram.

Perspective - Vanishing point - Estimation - Medicine - Health

Continue reading PrePrint: Equidistant Fish-Eye Calibration and Rectification by Vanishing Point Extraction

PrePrint: A Fast Bilinear Structure from Motion Algorithm Using a Video Sequence and Inertial ...

http://www.computer.org/tpami (IEEE transactions) — Fri, 20 Aug 2010 17:44:33 +0100

In this paper, we study the benefits of the availability of a specific form of additional information - the vertical direction (gravity) and the height of the camera both of which can be conveniently measured using inertial sensors, and a monocular video sequence for 3D urban modeling. We show that in the presence of this information, the SfM equations can be rewritten in a bilinear form. This allows us to derive a fast, robust, and scalable SfM algorithm for large scale applications. The SfM algorithm developed in this paper is experimentally demonstrated to have favorable properties compared to the sparse bundle adjustment algorithm. We provide experimental evidence indicating that the proposed algorithm converges in many cases to solutions with lower error than state-of-art implementations of bundle adjustment. We also demonstrate that for the case of large reconstruction problems, the proposed algorithm takes lesser time to reach its solution compared to bundle adjustment. We also present SfM results using our algorithm on the Google StreetView research dataset.

Algorithm - Google StreetView - Google - Gravitation - Research

Continue reading PrePrint: A Fast Bilinear Structure from Motion Algorithm Using a Video Sequence and Inertial Sensors

PrePrint: A Non-Rigid Kernel Based Framework for 2D3D Pose Estimation and 2D Image Segmentation

http://www.computer.org/tpami (IEEE transactions) — Fri, 20 Aug 2010 17:44:33 +0100

In this work, we present a non-rigid approach to jointly solve the tasks of 2D-3D pose estimation and 2D image segmentation. In general, most frameworks that couple both pose estimation and segmentation assume that one has the exact knowledge of the 3D object. However, this assumption may be violated if only a general class to which a given shape belongs is given (e.g., cars, boats, or planes). Thus, we propose to solve the 2D-3D pose estimation and 2D image segmentation via nonlinear manifold learning of 3D embedded shapes for a class of objects for which one may not be able to associate a skeleton model. Thus, the novelty of our method is threefold: Firstly, we derive a generalized gradient flow for the task of non-rigid pose estimation and segmentation. Secondly, due to possible nonlinear structures in one's training set, we evolve the pre-image obtained through kernel PCA for the task of shape analysis. Thirdly, we show that the derivation for shape weights is general. This allows us to use various kernels with only minimal changes needed to be made to the overall evolution scheme. We provide experimental results that highlight the algorithm's robustness on challenging pose estimation and segmentation scenarios.

Linux - Operating Systems - Kernel - OpenSUSE - Linux kernel

Continue reading PrePrint: A Non-Rigid Kernel Based Framework for 2D3D Pose Estimation and 2D Image Segmentation

PrePrint: Applying Property Testing to an Image Partitioning Problem

http://www.computer.org/tpami (IEEE transactions) — Fri, 20 Aug 2010 17:44:32 +0100

Property testing is a rapidly growing field of research. Typically, a property testing algorithm proceeds by quickly determining whether an input can satisfy some condition, under the assumption that most inputs do not satisfy it. If the input is "far" from satisfying the condition, the algorithm is guaranteed to reject it with high probability. Applying this paradigm to image detection is desirable, since images are large objects, and a lot of time can be saved by quickly rejecting images which are "far" from satisfying a certain condition the user is interested in. Further, typically most inputs are indeed "far" from the sought images. We demonstrate this by analyzing the problem of deciding whether a binary image can be partitioned according to a template represented by a rectangular grid, and introduce a quick "rejector" which tests an image extracted from the input image, but whose size, as well as the time required to construct it, are constants which are independent of the input image size. With high probability, the rejector dismisses the inputs which are "far" from the template.

Probability - Algorithm - Research - Business - Image Consultants

Continue reading PrePrint: Applying Property Testing to an Image Partitioning Problem

PrePrint: Multi-Feature Based High-Resolution Palmprint Recognition

http://www.computer.org/tpami (IEEE transactions) — Fri, 20 Aug 2010 17:44:32 +0100

Palmprint is promising biometric feature for use in access control and forensic applications. Previous research on palmprint recognition mainly concentrates on low resolution (about 100 ppi) palmprints. But for high-security applications (e.g. forensic usage), high resolution palmprints (500 ppi or higher) are required from which more useful information can be extracted. In this paper, we propose a novel recognition algorithm for high-resolution palmprint. The main contributions of the proposed algorithm include: (1) Use of multiple features, namely minutiae, density, orientation and principal lines for palmprint recognition to significantly improve the matching performance of the conventional algorithm. (2) Design of a quality based and adaptive orientation field estimation algorithm, which performs better than the existing algorithm in case of regions with large number of creases. (3) Use of a novel fusion scheme for identification application which performs better than conventional fusion methods, e.g. weighted sum rule, SVMs or Neyman-Pearson rule. Besides, we analyze the discriminative power of different feature combinations and find that density is very useful for palmprint recognition. Experimental results on the database containing 14,576 full palmprints show that the proposed algorithm has achieved a good performance.

Biometrics - Security - Access control - Research - Business

Continue reading PrePrint: Multi-Feature Based High-Resolution Palmprint Recognition

PrePrint: A Laplacian Approach to Multi-Oriented Text Detection in Video

http://www.computer.org/tpami (IEEE transactions) — Fri, 20 Aug 2010 17:44:31 +0100

In this paper, we propose a method based on the Laplacian in the frequency domain for video text detection. Unlike many other approaches which assume that text is horizontally-oriented, our method is able to handle text of arbitrary orientation. The input image is first filtered with Fourier-Laplacian. K-means clustering is then used to identify candidate text regions based on the maximum difference. The skeleton of each connected component helps to separate the different text strings from each other. Finally, text string straightness and edge density are used for false positive elimination. Experimental results show that the proposed method is able to handle graphics text and scene text of both horizontal and non-horizontal orientation.

String - Laplace operator - Image - Paper - Text

Continue reading PrePrint: A Laplacian Approach to Multi-Oriented Text Detection in Video

PrePrint: Space-time Super-Resolution Using Graph-cut Optimization

http://www.computer.org/tpami (IEEE transactions) — Fri, 20 Aug 2010 17:44:30 +0100

We address the problem of super-resolution -- obtaining high-resolution images and videos from multiple low resolution inputs. The increased resolution can be in spatial or temporal dimensions, or even in both.We present a unified framework which uses a generative model of the imaging process and can address spatial super-resolution, space-time super-resolution, image deconvolution, single image expansion, removal of noise and image restoration. We model a high resolution image or video as a Markov random field and use maximum a posteriori estimate as the final solution using graph-cut optimization technique. We derive insights in to what super-resolution magnification factors are possible and the conditions necessary for super-resolution. We demonstrate spatial super-resolution reconstruction results with magnifications higher than predicted limits of magnification. We also formulate a scheme for selective super-resolution reconstruction of videos to obtain simultaneous increase of resolutions in both spatial and temporal directions. We show that it is possible to achieve space-time magnification factors beyond what has been suggested in the literature by selectively applying super-resolution constraints. We present results on both synthetic and real input sequences.

Image resolution - Image processing - Image - Video - Magnification

Continue reading PrePrint: Space-time Super-Resolution Using Graph-cut Optimization

PrePrint: Contour Detection and Hierarchical Image Segmentation

http://www.computer.org/tpami (IEEE transactions) — Fri, 20 Aug 2010 17:44:29 +0100

This paper investigates two fundamental problems in computer vision: contour detection and image segmentation. We present state-of-the-art algorithms for both of these tasks. Our contour detector combines multiple local cues into a globalization framework based on spectral clustering. Our segmentation algorithm consists of generic machinery for transforming the output of any contour detector into a hierarchical region tree. In this manner, we reduce the problem of image segmentation to that of contour detection. Extensive experimental evaluation demonstrates that both our contour detection and segmentation methods significantly outperform competing algorithms. The automatically generated hierarchical segmentations can be interactively refined by user specified annotations. Computation at multiple image resolutions provides a means of coupling our system to recognition applications.

Algorithm - Computer vision - Cluster analysis - Segmentation - Paper

Continue reading PrePrint: Contour Detection and Hierarchical Image Segmentation

What if 'English Only' Isn't Wrong?

http://www.aaai.org/aitopics/html/current.html (AAAI) — Fri, 20 Aug 2010 00:00:00 +0100

Foreigners learn our language; we don't learn theirs. On the campaign trail in 2008, Barack Obama was asked about foreign-language education. Being able to speak a foreign language makes you "so much more employable," he said. " I recently telephoned Dennis Baron, professor of English and linguistics at the University of Illinois at Urbana-Champaign, and read him that quote. Americans are still stubbornly—even proudly—monolingual, more concerned with protecting English than with learning another tongue.

Continue reading What if 'English Only' Isn't Wrong?

Floor fight: Cleaning robot Mint versus Scooba

http://www.aaai.org/aitopics/html/current.html (AAAI) — Fri, 20 Aug 2010 00:00:00 +0100

Mint is a new floor-cleaning robot that can wipe down hard surfaces with dust cloths. How does it stack up to iRobot's Scooba? Newcomer Mint, at right, is smaller and lighter than iRobot's Scooba. But can it clean floors as well? But Evolution Robotics'Mintis one of a number of floor-cleaning robots designed to lighten the housework load. Mint is somewhat comparable to low-cost dusting robots like the $50RoboMop, but at $249 it's almost in the same league as more sophisticated floor-washing machines like iRobot'sScooba.

Continue reading Floor fight: Cleaning robot Mint versus Scooba