Posts tagged interaction
Articles and news from the latest research reports.
Robovie talking robot joins science class at Higashihikari Elementary School in Japan
Robovie a 1.2-meter robot developed by ATR joined the science class at Higashihikari Elementary School in Japan on Feb. 5 for the start of a 14-month experiment. Data will be gathered to improve the robot’s ability to interact naturally with multiple people. The robot has been given facial photos and voiceprints of 119 fifth graders and teachers. On the first day of class, Robovie greeted the students, and was asked by a teacher to answer what a “wound up copper wire” was. It answered, “A copper coil. It’s part of the motors that move my body.” During class Robovie waited at the back of the room, recognizing the faces of the students and recording their movements. After class it shook hands with sixth graders and answered their questions.
As part of research into the co-existence of humans and robots, the experiment with Robovie is being carried out at a school because the environment allows for the acquisition of large amounts of data from the movements of the children. The robot has been given facial photos and voiceprints of 119 fifth graders and teachers. Robovie’s daily conversation level is equivalent to a five-year-old human, but it has been programmed with the entire contents of a fifth-grade science textbook. This is the first experiment using a robot at a school to last over a year.
Functional Connectivity and Tuning Curves in Populations of Simultaneously Recorded Neurons
How interactions between neurons relate to tuned neural responses is a longstanding question in systems neuroscience. Here we use statistical modeling and simultaneous multi-electrode recordings to explore the relationship between these interactions and tuning curves in six different brain areas. We find that, in most cases, functional interactions between neurons provide an explanation of spiking that complements and, in some cases, surpasses the influence of canonical tuning curves. Modeling functional interactions improves both encoding and decoding accuracy by accounting for noise correlations and features of the external world that tuning curves fail to capture. In cortex, modeling coupling alone allows spikes to be predicted more accurately than tuning curve models based on external variables. These results suggest that statistical models of functional interactions between even relatively small numbers of neurons may provide a useful framework for examining neural coding.
Virtual reality ‘beaming’ technology transforms human-animal interaction
Using cutting-edge virtual reality technology, researchers have ‘beamed’ a person into a rat facility allowing the rat and human to interact with each other on the same scale.
Published in PLOS ONE, the research enables the rat to interact with a rat-sized robot controlled by a human participant in a different location. At the same time, the human participant (who is in a virtual environment) interacts with a human-sized avatar that is controlled by the movements of the distant rat. The authors hope the new technology will be used to study animal behaviour in a completely new way.
Computer scientists at UCL and the University of Barcelona have been working on the idea of ‘beaming’ for some time now, having last year digitally beamed a scientist in Barcelona to London to be interviewed by a journalist.
The researchers define ‘beaming’ as digitally transporting a representation of yourself to a distant place, where you can interact with the people there as if you were there. This is achieved through a combination of virtual reality and teleoperator systems. The visitor to the remote place (the destination) is represented there ideally by a physical robot.
According to new research, meerkats enhance their intelligence through nine different social and asocial mechanisms. What really makes these animals stand out is their intelligent coordinated behaviour, which rivals that of chimps, baboons, dolphins and even humans in its complexity and efficiency.
A team led by William Hoppitt of the University of St. Andrews presented wild meerkats with a novel foraging task to investigate the animal’s learning mechanisms. ‘The model deals with the rate at which individuals interact with the task, solve the task once they are interacting with it, or give up on the task when they are manipulating it,’ said Hoppitt.
They found that the meerkats engaged in a wide variety of social and asocial behaviours to learn to solve the task, and that in general the social factors helped draw the meerkats into the task, while the asocial processes helped them actually solve the task.
The model may also be more broadly applicable and can be used to investigate the relationship between social learning mechanisms and so-called ‘behavioural traditions’ that together can constitute a culture.
Physics and math shed new light on biology by mapping the landscape of evolution
August 8, 2012
Researchers capture evolutionary dynamics in a new theoretical framework that could help explain some of the mysteries of how and why species change over time.
Although the qualitative description of evolution – its observed behavior and characteristics – is well-established, a comprehensive quantitative theory that captures general evolution dynamics is still lacking. There are also many lingering mysteries surrounding the story of life on Earth, including the question of why sex is such a prevalent reproductive strategy. A team of scientists from the Chinese Academy of Sciences; Jilin University in Jilin, China; and the State University of New York at Stony Brook, led by Prof. Jin Wang, has examined some of these puzzles from a physical science prospective. They propose a new theory of evolution with two ingredients: the underlying emergent “fitness” landscape and an associated evolutionary force called “curl flux,” which causes species to move through the emergent fitness landscape in a spiraling manner.
The researchers captured evolutionary relationships in a system of equations. They then created quantitative pictures that visualized evolutionary pathways as journeys through a mountainous terrain of peaks and valleys of biological fitness. The key breakthrough beyond the conventional quantitative theory of evolution is the emergent curl flux, which is generated by interactions between individuals within or across species. The underlying emergent landscape gradient and the curl flux act together as a “Yin and Yang” duality pair to determine the dynamics of general evolution, says Wang. An example of similar behavior is the particle and wave duality that determines the dynamics of the quantum world, he notes. The researchers also note that this combined effect is analogous to the way electric and magnetic forces both act on electrons.
The new theory provides a physical foundation for general evolution dynamics. The researchers found that interactions between individuals of different species can give rise to the curl flux. This can sustain an endless evolution that does not lead to areas of higher relative fitness, even if the physical environment is unchanged.
This finding offers a theoretical framework to explain the Red Queen Hypothesis, which states that species continually evolve in order to fend off parasites that are themselves continually evolving. The hypothesis, first proposed by evolutionary biologist Leigh Van Valen in 1973, gets its name from the character of the Red Queen in Lewis Carroll’s book Through the Looking-Glass, who observed that in her world it was necessary to keep running just to stay in one place. The idea of endless co-evolution through the maintenance of the genetic variation due to the curl flux could help explain the benefits of sexual reproduction, since the mixing and matching of genes preserves a greater diversity of traits. When a species’ arms race with a co-evolving parasite takes an unexpected twist, a previously unnecessary trait could suddenly turn into the key to surviving. In the co-evolving world, there is no guarantee for “survival of the fittest” and it is often necessary to keep running for survival.
Source: PHYS.ORG