- UID
- 504111
- 在线时间
- 小时
- 注册时间
- 2010-1-19
- 最后登录
- 1970-1-1
- 主题
- 帖子
- 性别
- 保密
|
【计时1】
People Smile When They Are Frustrated, and the Computer Knows the Difference
ScienceDaily (May 28, 2012) — Do you smile when you're frustrated? Most people think they don't -- but they actually do, a new study from MIT has found. What's more, it turns out that computers programmed with the latest information from this research do a better job of differentiating smiles of delight and frustration than human observers do.
[attachimg=560,410]101575[/attachimg]
The research could pave the way for computers that better assess the emotional states of their users and respond accordingly. It could also help train those who have difficulty interpreting expressions, such as people with autism, to more accurately gauge the expressions they see.
"The goal is to help people with face-to-face communication," says Ehsan Hoque, a graduate student in the Affective Computing Group of MIT's Media Lab who is lead author of a paper just published in the IEEE Transactions on Affective Computing. Hoque's co-authors are Rosalind Picard, a professor of media arts and sciences, and Media Lab graduate student Daniel McDuff.
In experiments conducted at the Media Lab, people were first asked to act out expressions of delight or frustration, as webcams recorded their expressions. Then, they were either asked to fill out an online form designed to cause frustration or invited to watch a video designed to elicit a delighted response -- also while being recorded.
When asked to feign frustration, Hoque says, 90 percent of subjects did not smile. But when presented with a task that caused genuine frustration -- filling out a detailed online form, only to then find the information deleted after pressing the "submit" button -- 90 percent of them did smile, he says. Still images showed little difference between these frustrated smiles and the delighted smiles elicited by a video of a cute baby, but video analysis showed that the progression of the two kinds of smiles was quite different: Often, the happy smiles built up gradually, while frustrated smiles appeared quickly but faded fast.
In such experiments, researchers usually rely on acted expressions of emotion, Hoque says, which may provide misleading results. "The acted data was much easier to classify accurately" than the real responses, he says. But when trying to interpret images of real responses, people performed no better than chance, assessing these correctly only about 50 percent of the time.
【390】
【计时2】
Understanding the subtleties that reveal underlying emotions is a major goal of this research, Hoque says. " eople with autism are taught that a smile means someone is happy," he says, but research shows that it's not that simple.
While people may not know exactly what cues they are responding to, timing does have a lot to do with how people interpret expressions, he says, For example, former British prime minister Gordon Brown was widely seen as having a phony smile, largely because of the unnatural timing of his grin, Hoque says. Similarly, a campaign commercial for former presidential candidate Herman Cain featured a smile that developed so slowly -- it took nine seconds to appear -- that it was widely parodied, including a spoof by comedian Stephen Colbert. "Getting the timing right is very crucial if you want to be perceived as sincere and genuine with your smiles," Hoque says.
Jeffrey Cohn, a professor of psychology at the University of Pittsburgh who was not involved in this research, says this work "breaks new ground with its focus on frustration, a fundamental human experience. While pain researchers have identified smiling in the context of expressions of pain, the MIT group may be the first to implicate smiles in expressions of negative emotion."
Cohn adds, "This is very exciting work in computational behavioral science that integrates psychology, computer vision, speech processing and machine learning to generate new knowledge … with clinical implications." He says this "is an important reminder that not all smiles are positive. There has been a tendency to 'read' enjoyment whenever smiles are found. For human-computer interaction, among other fields and applications, a more nuanced view is needed."
In addition to providing training for people who have difficulty with expressions, the findings may be of interest to marketers, Hoque says. "Just because a customer is smiling, that doesn't necessarily mean they're satisfied," he says. And knowing the difference could be important in gauging how best to respond to the customer, he says: "The underlying meaning behind the smile is crucial."
The analysis could also be useful in creating computers that respond in ways appropriate to the moods of their users. One goal of the research of Affective Computing Group is to "make a computer that's more intelligent and respectful," Hoque says.
【382】
【计时3】
Will a Volcano Erupt? The Answer's in the Crystals
by Erin Loury on 24 May 2012, 3:01 PM
[attachimg=800,600]101576[/attachimg]
Active volcanoes often send out signals advertising that they are awake: small earthquakes and venting gasses usually aren't good news. But often, the messages aren't clear. Now, researchers have another tool to help predict when a volcano may blow. A new study shows that chemical patterns in volcanic crystals match up with patterns in volcanic earthquake and gas recordings—giving scientists a chance to save thousands of lives before it's too late.
Tiny volcanic crystals, often just 50 to 100 micrometers across, float suspended in magma, the ultrahot mix of molten rock and dissolved gases that can rise beneath volcanoes. Many of these crystals have concentric bands that look like tree rings. Events like a new pulse of hot magma into the chamber— like that which can precede an eruption—can cause elements inside a crystal, such as iron and magnesium, to migrate toward the crystal's core or toward its edges, creating new rings. Once a minor eruption preceding the main event spews a crystal-bearing magma above ground, it all solidifies and locks in the record of the volcano's past, which geologists studying the threat from the volcano can collect and interpret.
Kate Saunders, a volcanologist at the University of Bristol in the United Kingdom, wondered whether the crystals could also be used to predict the future. She and colleagues studied a mineral crystal called orthopyroxene from Mount St. Helens in Washington. The active volcano erupted catastrophically on 18 May 1980, and continued to produce smaller eruptions through October of 1986. The team characterized the chemical patterns of 98 crystals collected over the course of the eruptions and compared those patterns with records of earthquakes and gas release that other researchers collected during this same time period.
【305】
【计时4】
Crystals start to build up a year prior to eruptions and peak just before an explosion happens, the team reports online today in Science. Crystals with magnesium-rich rims and iron-rich cores, which signify heating from the intrusion of new magma, were associated with deep earthquakes. A spike in these magnesium-rimmed crystals also occurred just prior to the massive 18 May eruption, indicating that pulses of new magma preceded the blast. Crystals with iron-rich rims and magnesium-rich cores, which form when the magma is degassing and cooling, corresponded with peaks in sulfur dioxide gas release. These crystals peaked prior to later eruptions at Mount St. Helens.
Knowing how the chemical fingerprints of crystals link up with other recorded signals will help scientists read a volcano's past to better interpret its future warning signs, says Saunders. "We can tell if we expect to see new pulses of magma, or if we expect the magma to sit there and degas, and we can start to work out what signs we should be looking for in the monitoring data" to predict eruptions.
"It's a neat package, the fact that they can work backwards with these crystals to nail down the timing of magmatic [activity]," says Carl Thornber, a volcanologist with the United States Geological Survey at the Cascades Volcano Observatory in Vancouver, Washington. "It gives us much more solid information to interpret what's going on down below, and how to interpret all the measurements we're making."
"The technique can be transferred to any volcano," Saunders says. Scientists can't monitor the moving and shaking of every volcano worldwide, but they can often collect the products of lesser eruptions, such as these deep-formed crystals, to understand the volcano's behavior. They can also watch for spikes in the types of crystals that might foretell an explosion. "If we look at these erupted products, we can build up a picture of what's happened and what we expect to happen if the volcano suddenly reawakes."
【326】
【计时5】
How Many Stars Never Make the Big Time?
by Ken Croswell on 24 May 2012, 4:05 PM
[attachimg=625,600]101577[/attachimg]
Becoming a star can be a challenge. But new observations reveal that it's much easier in space than in Hollywood. Only about 14% of all aspiring celestial stars fizzle out, researchers report.
In principle, it's easy to deduce how many stars succeed: just compare the numbers of normal stars with the number of failed stars, also known as brown dwarfs. These flops are born with less than 8% of the sun's mass, so their centers never heat up enough to sustain the nuclear fusion of hydrogen-1, the isotope that powers so-called main-sequence stars like the sun. Most brown dwarfs do burn hydrogen-2, or deuterium, but it soon runs out, and all nuclear reactions cease. When young, a brown dwarf glows red—chiefly from the heat of its birth—and then cools and fades as it ages. That makes them hard to find—and thus hard to know whether they're rare or as plentiful as full-fledged stars.
Astronomer J. Davy Kirkpatrick of the California Institute of Technology in Pasadena and colleagues decided to take a look at infrared wavelengths, where the objects emit most of their radiation. The team used NASA's Wide-field Infrared Survey Explorer (WISE) spacecraft, launched in late 2009, to detect brown dwarfs near the sun, some of which have cooled to room temperature. WISE spotted 16 previously unknown brown dwarfs within 26 light-years of Earth. As the astronomers will report in the 10 July issue of The Astrophysical Journal, comparing the number of brown dwarfs with the number of full-fledged stars suggests that the solar neighborhood has one failed star for every six success stories.
【284】
【自由】
"It was surprising to me," says Kirkpatrick, who had expected to find far more brown dwarfs. If they were more common, one might reside even closer than Alpha Centauri, the nearest star system to the sun. "It's a lot less likely we're going to find any brown dwarf that's that close." The two nearest known brown dwarfs orbit Epsilon Indi, a star 11.8 light-years away, which is nearly three times farther from Earth than Alpha Centauri.
Astronomer Aleks Scholz of the Dublin Institute for Advanced Studies says the new estimate for the number of brown dwarfs agrees with his team's result from a very different approach: searching star clusters so young their brown dwarfs haven't had time to fade, making them easier to spot. He and other astronomers have found an approximately 1:5 ratio between brown dwarfs and normal stars.
However, astronomer Todd Henry of Georgia State University in Atlanta is critical of the new work's conclusion. "I think it's premature," says Henry, who suspects that many of the brown dwarfs WISE has found are farther away from Earth than claimed. That would mean they are less abundant than the researchers estimated, and Henry says the ratio of brown dwarfs to normal stars could be anywhere between 1:5 and 1:20.
【284】
【越障】
New Genetic Method Developed to Pinpoint Individuals' Geographic Origin
ScienceDaily (May 24, 2012) — Understanding the genetic diversity within and between populations has important implications for studies of human disease and evolution. This includes identifying associations between genetic variants and disease, detecting genomic regions that have undergone positive selection and highlighting interesting aspects of human population history.
[attachimg=560,376]101578[/attachimg]
Now, a team of researchers from the UCLA Henry Samueli School of Engineering and Applied Science, UCLA's Department of Ecology and Evolutionary Biology and Israel's Tel Aviv University has developed an innovative approach to the study of genetic diversity called spatial ancestry analysis (SPA), which allows for the modeling of genetic variation in two- or three-dimensional space.
Their study is published online this week in the journal Nature Genetics.
With SPA, researchers can model the spatial distribution of each genetic variant by assigning a genetic variant's frequency as a continuous function in geographic space. By doing this, they show that the explicit modeling of the genetic variant frequency -- the proportion of individuals who carry a specific variant -- allows individuals to be localized on a world map on the basis of their genetic information alone.
"If we know from where each individual in our study originated, what we observe is that some variation is more common in one part of the world and less common in another part of the world," said Eleazar Eskin, an associate professor of computer science at UCLA Engineering. "How common these variants are in a specific location changes gradually as the location changes.
"In this study, we think of the frequency of variation as being defined by a specific location. This gives us a different way to think about populations, which are usually thought of as being discrete. Instead, we think about the variant frequencies changing in different locations. If you think about a person's ancestry, it is no longer about being from a specific population -- but instead, each person's ancestry is defined by the location they're from. Now ancestry is a continuum."
The team reports the development of a simple probabilistic model for the spatial structure of genetic variation, with which they model how the frequency of each genetic variant changes as a function of the location of the individual in geographic space (where the gene frequency is actually a function of the x and y coordinates of an individual on a map).
"If the location of an individual is unknown, our model can actually infer geographic origins for each individual using only their genetic data with surprising accuracy," said Wen-Yun Yang, a UCLA computer science graduate student.
"The model makes it possible to infer the geographic ancestry of an individual's parents, even if those parents differ in ancestry. Existing approaches falter when it comes to this task," said UCLA's John Novembre, an assistant professor in the department of ecology and evolution.
SPA is also able to model genetic variation on a globe.
"We are able to also show how to predict the spatial structure of worldwide populations," said Eskin, who also holds a joint appointment in the department of human genetics at the David Geffen School of Medicine at UCLA. "In just taking genetic information from populations from all over the world, we're able to reconstruct the topology of the global populations only from their genetic information."
Using the framework, SPA can also identify loci showing extreme patterns of spatial differentiation.
"These dramatic changes in the frequency of the variants potentially could be due to natural selection," Eskin said. "It could be that something in the environment is different in different locations. Let's say a mutation arose that has some advantageous property in a certain environment. So you can imagine then that a kind of force for genetic selection would make this mutation more common in that environment."
The research team began to examine all of the genes, and for each gene they computed how sharp of a change there was in the frequencies. They soon discovered that the genes which had the largest and most extreme changes are the ones that are known to have experienced selection in the recent past.
"So this is a new method for finding genes that are also undergoing selection in humans," Yang said.
【705】
|
本帖子中包含更多资源
您需要 登录 才可以下载或查看，没有帐号？立即注册
x
|
京公网安备11010202008513号 京ICP证101109号 京ICP备12012021号
发表于 2012-5-29 23:56:58
收藏
收藏
辛苦啦