Skip Navigation

Duke Research - February 2011

Alex Trebek in parade

February 23, 2011

Training Machines to Recognize People

We've established that a computer can beat the best humans at Jeopardy.  But can it tell Alex Trebek from Chuck Woolery?

On Feb. 21, Deva Ramanan, Assistant Professor of Computer Science at UC Irvine, was at Duke talking about his work on enhancing computer vision to the point where machines might be able to recognize individual people.

He said it has been extremely difficult for a computer to recognize humans in different images or video sequences because of significant variations in every frame of reference.  “Finding people in images is difficult [because of] variation in illumination, appearance, pose, viewpoint and background clutter. These are the classic nuisance factors for object recognition,” Ramanan said.

Ramanan and his team have worked on adaptive and dynamic algorithms  to solve this problem by breaking down the entire image template into local, global and temporal models.

They also extend the template into semantic parts, which implies that the computer program is already told a priori the parts of a body that it needs to learn. The computer then has to learn the different appearances and the template based on the training data set.

“One can also apply suppression techniques—for example, two objects can never occupy the same 3D volume, people don’t stand on top of each other but stand next to each other, and bottles are often supported by tables.” Using a variety of models, the team has been able to build a relatively robust model of human-detection in images and movies.

Applications of such a people-detection system include video surveillance, autonomous vehicle navigation, healthcare, image and movie search and building smarter visual interfaces.

Ramanan is the recipient of the 2009 David Marr Prize, an NSF Career Award and the 2010 PASCAL Visual Object Class Challenge  Lifetime Achievement Award.

Posted by vm19. 0 comments

Tags: computers/technology, lecture

Ben Wildavsky

February 22, 2011

A Global Economy of Scholarship

“In this world of global mobility, people can get ahead more based on what they know, not just who they are,” author Ben Wildavsky said in a Feb. 21 lecture as part of the "Re-imagining the Academy" series. 

“The best students in the world can really write their own tickets.”

Wildavsky is a senior fellow at the Kauffman Foundation and previously edited America's Best Colleges, published by U.S. News & World Report. His lecture, sponsored by the Bass Society of Fellows, Duke Libraries, the Office of Global Strategy & Programs and the Office of Public Affairs & Government Relations, was the third in the annual series.

Wildavsky shared three important trends that he discovered while writing "The Great Brain Race: How Global Universities are Reshaping the World,” a book about international higher education.

1. Unprecedented academic mobility
There are 3 million globally mobile students in the world -- students who are willing to travel out of their home countries to pursue an education. According to Wildavsky, this number has increased 57% in a decade, and by all estimates will climb further still. For confirmation, one only needs to look at Duke’s international applicant pool. Wildavsky said  10 Chinese students sought admission in 2004, while 730 applied to be part of the class of 2014. And students are not the only mobile ones; faculty members are increasingly mobile as well.

2.    Emergence of global college rankings
Ranked lists of American schools have spurred rankings that examine universities around the world. (Out of 500+ institutions, The Times (UK) ranked Duke 24 and the Academic Ranking of World Universities put Duke at 35.) Wildavsky admitted that ranking systems are controversial and can have flaws. However, he believes that they’re still useful because they provide an external yard stick and “can help universities improve when they’re not measuring up as much as they’d like to.”

3.    Race to create world-class institutions
Many countries don’t just want to send their students abroad or establish branch campuses of international universities; they want to create their own world-class institutions. This can be done by allocating more funds to the most competitive universities, as the French did. Of course, it also helps to have a generous benefactor: King Abdullah of Saudi Arabia reportedly gave billions of dollars to establish the King Abdullah University of Science and Technology.

According to Wildavsky, the transition to global education is a huge opportunity to foster talent and innovation. However, not everyone is excited. He cited two main concerns:

1.    Brain drain, whereby promising students leave, taking the skills they learned with them

2.    Talented foreigners crowding out domestic students.

“There’s this sense that the Chinese are going to eat our lunch,” Wildavsky said.

He proposed the term “brain circulation” instead of “brain drain,” explaining that talented students and faculty often follow opportunities across the globe, and many times choose to return home again. As for the influx of talented international students, he argued that many programs have just expanded to support them.

“I think the theory is that, if others are getting ahead, we’re falling behind,” Wildavsky said. “But we need well-trained minds, whatever their nationality, to solve the world’s great problems.” He argued that innovation and economic growth all depend on the most free movement of people and ideas.

“Knowledge is not a finite resource like gold or diamonds. It’s something that can grow.”

Posted by rlb35. 0 comments

Tags: faculty, lecture, students

Anopheles gambiae from http://dx.doi.org/10.1371/journal.pbio.0040246

February 14, 2011

Statistical Research on Health Care in Kenya

“Malaria counts for 40 percent of the hospital visits in most areas of Kenya,” Duke Global Health Institute  researcher Nathan Smith said during a Friday, Feb. 11 talk.

Nathan’s research looks at the knowledge levels and the available preventive measures for malaria in different districts of Kenya. His most recent project focuses on households in the Bungoma district in Western Kenya. By using GPS data and the AMPATH home-based counseling and testing program, a complete census for that area was first created. Further data was collected to determine which of the households had at least one malaria-resistant bednet.

Nathan noted that a lot of household and individual factors had to be taken into account for a data regression model to be created. “We conducted data analysis, density, distance and cluster analysis, as well as logistic regression  to arrive at the results.”

His research shows that 79 percent of the households did not own a bednet. Another interesting result is that 95-98 percent of the children in these households had BCG immunization, however quite a few were missing their measles immunization.

To further make sense of the data, Nathan and his team calculated the coverage percentage. The measurements show that households with children and rural households were more likely to have at least one bednet, compared to urban households and households without children.

Another problem that Nathan highlighted was that neighborhood chemists create their own mixture of medicine that they think the patients need at that point of time.

“One of the major concerns is the unregulated use of medicine to treat malaria in these areas.” This is a complex issue since the average distance from a household to public health facilities is almost double compared to these retail outlets. “Moving forward, we want to characterize the type of care in terms of drug availability, diagnostic capacity, cost to the patient, dispensing practices and provider knowledge.”

Posted by vm19. 0 comments

Tags: global health, lecture, medicine, research

February 11, 2011

Building, Not Just Re-Building

Thirty-four seconds doesn’t seem like a lot of time, but on January 12, 2010 that was all it took for a 7.8-scale earthquake to kill 230,000 people in Haiti. The quake destroyed 28 out of 29 government ministries, crippling the government’s ability to respond.

Fortunately, international aid was quick to arrive. Donations flooded in at an unprecedented scale and countries offered their support. Within 24 hours, the United States Agency for International Development (USAID) had an emergency response team on the ground.

“The first thing we had to focus on was the rubble,” said Russell Porter coordinator of the USAID long-term Haiti Task Team.  “I don’t know how to describe what that looked like," he said. "If you stacked dump trucks, they’d go a third of the way around the world. You’ve got to take that stuff away before you can rebuild.”

Porter shared some of his experiences before and after the quake in the Global Health Institute’s Fourth-Annual Lecture on Global Health, Feb. 3, which was also part of A World Together.

The quake left 1.5 million people homeless, and USAID was tasked with finding safe shelter for them. One tactic was to assess buildings that hadn’t fallen down; if they were safe, people could move back in. That, in addition to construction of new shelters, helped bring the number of homeless people down to 810,000.

“That’s a big drop, but it’s still a lot of people,” Porter said. “The next 810,000 -- that’s going to be more of a challenge.”

Even before the earthquake, Haiti needed assistance, ranking 145/169 on the UN Human Development Index (compared to its neighbor, the Dominican Republic, at 88). “We’re not just rebuilding Haiti, we’re building it,” Porter said. He cited food security, infrastructure and health as priorities for long-term development.

Several months into the recovery effort, this progress was palpable. But, “then there was cholera -- as if Haiti didn’t have enough!”

Almost 4,000 people have died since the outbreak began in October. Haiti has notoriously poor water quality, and residents frequently suffer from diarrhea. The danger, Porter said, is that cholera can kill within a day. If Haitians don’t know they have cholera, they cannot receive life-saving treatment.

USAID has launched several education campaigns that describe symptoms and tout the effectiveness of treating water with chlorine. Public awareness is spreading, but Porter still expects the outbreak to persist “for the foreseeable future.” 

Unfortunately, the cholera outbreak is sucking resources away from other recovery programs, and USAID can’t divert money from elsewhere to cover the shortfall. “We’re not taking money away from AIDS, we’re not taking money away from malaria,” Porter said. “It’s a real balancing act.”

Posted by rlb35. 0 comments

Tags: global health, lecture

history of autism graphic

February 9, 2011

Tigers and helicopters and refrigerators, oh mom!

Mothers can be tigers, helicopters or soccer moms. But no stereotype was perhaps ever as biting as “refrigerator mom”.

Beginning in the 1950’s, psychoanalysts blamed these mothers for their children's autistic and schizophrenic behavior, including rigid rituals, speech difficulty and self-isolation. Duke pediatrician and historian Jeffrey Baker reviewed the sad history of this belief at a Sociology-Psychology brown bag seminar on Feb. 8.

Psychiatrist Leo Kanner first described  autism in 1943 and suspected a neurological origin. Yet, nearly all of the parents, particularly the fathers, of the eleven autistic children he saw were highly intelligent, self-absorbed in their careers and emotionally aloof. They kept their children “neatly in a refrigerator that did not defrost,” he wrote in 1949.

Kanner focused specifically on his patients’ fathers. But Bruno Bettelheim, one of the greater “villains” in this history of autism, found the greatest fault with patients' mothers. 

Bettelheim shaped this “refrigerator mother paradigm” through his popular articles, like “Joey: A 'Mechanical Boy,” which appeared in Scientific American, and books, like The Empty Fortress. The theoretical psychoanalyst blamed mothers' subconscious neglect for the children's autism, and he refused to let others review his work.

He had no systematic way of tracking his patients’ progress, and he found ways of making anomalous data fit within the refrigerator mother framework, which became persuasive to so many people, Baker said.

The mothers were finally vindicated in the late seventies and early eighties, but this story perfectly illustrates an erroneous scientific paradigm and is an example of what scientists should watch for in their current research, he said.

 

Posted by ay37. 0 comments

Tags: behavior/psychology, faculty, lecture, medicine, neuroscience, science communication&education

subscribe

Keep me posted on research news. Learn more >>

RSS iconRSS

All fields required