Pandemic
-
Deep Learning-Based Cough Recognition Model Helps Detect the Location of Coughing Sounds in Real Time
The Center for Noise and Vibration Control at KAIST announced that their coughing detection camera recognizes where coughing happens, visualizing the locations. The resulting cough recognition camera can track and record information about the person who coughed, their location, and the number of coughs on a real-time basis.
Professor Yong-Hwa Park from the Department of Mechanical Engineering developed a deep learning-based cough recognition model to classify a coughing sound in real time. The coughing event classification model is combined with a sound camera that visualizes their locations in public places. The research team said they achieved a best test accuracy of 87.4 %.
Professor Park said that it will be useful medical equipment during epidemics in public places such as schools, offices, and restaurants, and to constantly monitor patients’ conditions in a hospital room.
Fever and coughing are the most relevant respiratory disease symptoms, among which fever can be recognized remotely using thermal cameras. This new technology is expected to be very helpful for detecting epidemic transmissions in a non-contact way. The cough event classification model is combined with a sound camera that visualizes the cough event and indicates the location in the video image.
To develop a cough recognition model, a supervised learning was conducted with a convolutional neural network (CNN). The model performs binary classification with an input of a one-second sound profile feature, generating output to be either a cough event or something else.
In the training and evaluation, various datasets were collected from Audioset, DEMAND, ETSI, and TIMIT. Coughing and others sounds were extracted from Audioset, and the rest of the datasets were used as background noises for data augmentation so that this model could be generalized for various background noises in public places.
The dataset was augmented by mixing coughing sounds and other sounds from Audioset and background noises with the ratio of 0.15 to 0.75, then the overall volume was adjusted to 0.25 to 1.0 times to generalize the model for various distances.
The training and evaluation datasets were constructed by dividing the augmented dataset by 9:1, and the test dataset was recorded separately in a real office environment.
In the optimization procedure of the network model, training was conducted with various combinations of five acoustic features including spectrogram, Mel-scaled spectrogram and Mel-frequency cepstrum coefficients with seven optimizers. The performance of each combination was compared with the test dataset. The best test accuracy of 87.4% was achieved with Mel-scaled Spectrogram as the acoustic feature and ASGD as the optimizer.
The trained cough recognition model was combined with a sound camera. The sound camera is composed of a microphone array and a camera module. A beamforming process is applied to a collected set of acoustic data to find out the direction of incoming sound source. The integrated cough recognition model determines whether the sound is cough or not. If it is, the location of cough is visualized as a contour image with a ‘cough’ label at the location of the coughing sound source in a video image.
A pilot test of the cough recognition camera in an office environment shows that it successfully distinguishes cough events and other events even in a noisy environment. In addition, it can track the location of the person who coughed and count the number of coughs in real time. The performance will be improved further with additional training data obtained from other real environments such as hospitals and classrooms.
Professor Park said, “In a pandemic situation like we are experiencing with COVID-19, a cough detection camera can contribute to the prevention and early detection of epidemics in public places. Especially when applied to a hospital room, the patient's condition can be tracked 24 hours a day and support more accurate diagnoses while reducing the effort of the medical staff."
This study was conducted in collaboration with SM Instruments Inc.
Profile: Yong-Hwa Park, Ph.D.
Associate Professor
yhpark@kaist.ac.kr
http://human.kaist.ac.kr/
Human-Machine Interaction Laboratory (HuMaN Lab.)
Department of Mechanical Engineering (ME)
Korea Advanced Institute of Science and Technology (KAIST)
https://www.kaist.ac.kr/en/
Daejeon 34141, Korea
Profile: Gyeong Tae Lee
PhD Candidate
hansaram@kaist.ac.kr
HuMaN Lab., ME, KAIST
Profile: Seong Hu Kim
PhD Candidate
tjdgnkim@kaist.ac.kr
HuMaN Lab., ME, KAIST
Profile: Hyeonuk Nam
PhD Candidate
frednam@kaist.ac.kr
HuMaN Lab., ME, KAIST
Profile: Young-Key Kim
CEO
sales@smins.co.kr
http://en.smins.co.kr/
SM Instruments Inc.
Daejeon 34109, Korea
(END)
2020.08.13 View 13675 -
Education, a Silver Lining in the Dark COVID-19 Cloud
If there is a silver lining behind the COVID-19 pandemic clouds engulfing the world in darkness, it would be ‘education’. The disruption caused by the pandemic has reminded us of the skills that students need in this unpredictable world and raised public awareness of guaranteeing continuous, fair, and quality learning opportunities. Educational innovation can become a positive and powerful catalyst to transform the world for a better future in the post-COVID era.
According to the speakers at the virtual forum co-hosted by the Global Strategy Institute (GSI) and Korea Policy Center for the Fourth Industrial Revolution (KPC4IR) at KAIST on June 24, the recent transition to remote education amplifies the existing socio-economic disparities between the haves and the have-nots, and narrowing the digital divide is the most urgent challenge that should be addressed in this ever-evolving technology-dominating era. They also called for students to be resilient despite the numerous uncertainties ahead of them and prepare new skill sets to better adjust to new environments.
KAIST launched the GSI as its think tank in February of this year. The GSI aims to identify global issues proactively and help make breakthroughs well aligned with solid science and technology-based policies. The second forum of the KAIST GSI, following its inaugural forum in April, was held under the theme “Envisioning the Future of Education for a Non-Contact Society in the Post-Coronavirus Era”.
In his opening remarks, KAIST President Sung-Chul Shin stressed that “distance teaching and learning will eventually become integral components of our future education system”. He then called for close collaboration between the public and private sectors to better shape the future of digital education. President Shin said that global cooperation is also needed to continue offering inclusive, quality education that can equally benefit every student around the world. “We should never let a crisis go to waste, and the COVID-19 pandemic is no exception,” he added.
CEO of Minerva Schools Ben Nelson described the current coronavirus crisis as “an earthquake happening deep down on the ocean floor – we don’t feel it, but it can cause a devastating tsunami.” He continued, “Online learning can totally change the current education system forever.”
Saying that blended education, which combines online and offline classes, will be the new norm in the post-coronavirus era, Coursera CEO Jeff Maggioncalda anticipates that institutions will have to offer more and more online courses and credentials, and should at the same time prepare to drive down the cost of education as students expect to pay much less in tuition and fees for online learning options. “With the economy slumping and unemployment soaring, job-relevant education will also be a must,” Maggioncalda said.
National University of Singapore President Tan Eng Chye further pointed out that future education systems should prepare students to be creative lifelong learners. President Tan encouraged students to be able to integrate knowledge and technical skills from multiple disciplines for complex problem solving, and be adaptable and resilient with bigger appetites for risks and a higher tolerance for failures. He also mentioned digital competency, empathy, and social responsibility as virtues that students in the post-coronavirus era should possess.
Rebecca Winthrop, Co-Director of the Center for Universal Education at the Brookings Institution, raised concerns over the ever-growing digital disparities caused by the recent shift to online teaching and learning, claiming that insufficient infrastructures for low-income families in developing nations are already causing added educational disparities and provoking the inequity issue around the world. “New approaches to leapfrog inequality and provide quality education equally through faster and more effective means should be studied,” she said.
In response to this, Vice President of Microsoft Anthony Salcito introduced the Microsoft Education Transformation Framework, which provides practical advice to develop strategies for digital education transformation with a holistic, long-term view implemented in discrete phases that the global community can begin today. The Framework reportedly shows how emerging technologies, such as artificial intelligence, support new approaches to building efficient and effective physical and digital infrastructure, modernizing teaching and learning, empowering research, and managing student success.
The GSI will host two more forums in September and November.
(END)
2020.06.24 View 11854 -
KAIST Forum Envisions Education in the Post-Covid Era
Global leaders including the CEOs of Minerva and Coursera to join the KAIST online forum to discuss how to facilitate inclusive educational environment amidst the ever-growing digital disparities
An international forum hosted by the KAIST Global Strategy Institute will examine how the disruptions caused by the global pandemic will impact the future of education. Global leaders will reflect on ways to better facilitate inclusive educational environments and mitigate the digital divide, especially in an era where non-contact environments are so critical.
The online forum to be held on June 24 from 09:00 am KST will livestream on YouTube and KTV. This is the second forum hosted by the GSI following its inaugural forum in April.
Minerva School’s CEO Ben Nelson and Coursera CEO Jeff Maggioncalda will be among the 15 speakers who will share their insights on the new transformations in the education sector.
The digital transformation of higher education will be the key topic every speaker will highlight to predict the future education in the post-Covid era. According to UNESCO and UNICEF, 1.6 billion students from 192 countries, which account for 91 percent of the student population in the world, have experienced educational disruptions in the past four months. Approximately 29 percent of the youth worldwide, around 346 million individuals, are not online.
KAIST President Sung-Chul Shin’s opening remarks will stress that technological breakthroughs should be used to benefit us all and the private and public sectors should collaborate to facilitate an inclusive educational environment.
Ben Nelson believes that global universities are at the point of inflection for making tough choices to reform higher education. He will introduce what will affect the decision-making procedure for investing in the digital transformation and the best recipe for building a successful remote learning environment.
Dr. Paul Kim, CTO and Assistant Dean of Stanford Graduate School of Education, will analyze the ramifications brought about by Covid-19 among both advanced countries and developing countries, and propose an optimal educational model for developing countries.
Phil Baty, Chief Knowledge Officer at Times Higher Education, will present the key survey results the Times Higher Education made with approximately 200 university presidents on how higher education will adapt in the years to come.
As for innovation in higher education, Vice President at Microsoft Anthony Salcito and Professor Tae Eog Lee from the Department of Industrial and Systems Engineering at KAIST will discuss the education innovation solutions they are currently working on and how their projects will continue to develop.
National University of Singapore President Tan Eng Chye will also opine on how education could be more accessible. He will share what is exacerbating educational inequity and how to ensure an inclusive learning environment.
The second session will cover how to cope with the digital inequity.
Director General at the Ministry of Science and ICT Sang Wook Kang will explain the unavoidable online transition that is required to address the educational disruptions. He will also share his ideas on how this crisis can be leveraged to advance the educational environment.
Meanwhile, Rebecca Winthrop, senior fellow and co-director for universal education at Brooking Institution, and Sooinn Lee, CEO and Creative Lead of Enuma, will present on how to reduce the educational disparity during the un-contact era.
Director Joung-Ho Kim at the GSI, who is the organizer of the forum, said that KAIST has been the forerunner in the educational innovation. He hopes that this online forum will provide meaningful momentum to reshape the future of education by addressing the challenges and disruptions this pandemic has caused.
URL Link to Live-Streaming Service: https://www.youtube.com/c/KAISTofficial
2020.06.22 View 9917 -
Antivirus Industry the Centerpiece of New Deal R&D Initiatives
- KAIST launches post-COVID-19 R&D initiatives for smart mobile medical systems. -
KAIST will make the antivirus industry the centerpiece of what it is touting as the KAIST New Deal R&D initiative, which will drive new growth engines for preparing for the post-coronavirus era.
According to the new initiative, KAIST will concentrate on creating antivirus technologies, infectious disease-related big data management, and non-contact services platforms as key future R&D projects.
President Sung-Chul Shin launched the COVID-19 R&D Initiative task force last month, composed of more than 50 professors from the Graduate School of Medical Science and Engineering, the Department of Biological Sciences, the College of Engineering, and the Department of Industrial Design. The task force came up with key research agendas that will promote smart mobile medical systems in the years ahead.
“We will devote all of our R&D capacities to pursue a smart healthcare society,” said President Shin. “Our competitiveness in the fields of AI, ICT, materials, and bio-technology holds significant potential for building a healthy society powered by smart medical systems in Korea,” he added.
The smart medical systems focus mainly on building an Epidemic Mitigating Mobile Module (EMMM). The EMMM will manage epidemics via the three phases of prevention, emergency response, and treatment, with the development of each phase’s technological modules. The EMMM will also build an AI big data platform to assist with clinical applications and epidemic management.
Technologies applicable for the prevention phase include developing recyclable antivirus masks, plasma virus sterilizers, and smart breathable protective gowns. KAIST researchers will also focus on developing diagnosis modules that will identify epidemics more quickly and accurately.
Most significantly, KAIST aims to develop technologies for anti-infection medical services such as the transformable negative pressure ambulance module and negative pressure room, which are specially developed for respiratory infections.
The new R&D initiatives will center on virus therapies and treatments, specifically pushing forward vaccine and robotics studies. As caring robots and delivery robots will become common as main caregivers via noncontact services, research focusing on robotics will be significantly enhanced.
Even before launching the new R&D initiatives, researchers have started to present new technologies to help address the pandemic. Professor Il-Doo Kim’s team in the Department of Materials Science and Engineering developed a washable nano-fiber filtered face mask that is preparing for commercialization.
GPS tracking of infections has expanded comprehensively to detect both indoor and outdoor activities of infected patients. Professor Dong-Soo Han from the School of Computing developed Wi-Fi positioning software built into mobile phones that can trace both activities and is now preparing to roll it out.
Virologist Ui-Cheol Shin from the Graduate School of Medical Science and Engineering is carrying out research on a universal T-cell vaccine that can block the Betacoronaviruses. It is reported that that new epidemics such as SARS, MERS, and COVID-19 carry Betacoronaviruses.
Research teams in the Graduate School of AI are conducting various research projects on building prediction models for outbreaks and spreads using big data.
(END)
2020.05.20 View 10641 -
Long Economic Depressions and Disparities Loom in the Wake of the COVID-19
"Global Cooperation for Managing Data Key to Mitigating the Impacts Around the World"
<Full recorded video of the GSI-IF2020>
The COVID-19 pandemic will lead to long economic depressions around the entire world. Experts predicted that the prevalent inequities among the countries, regions, and individuals will aggravate the economic crisis. However, crises always come with new opportunities and international cooperation and solidarity will help creating a new normal in the post-coronavirus era.
In a very basic but urgent step, global cooperation for managing data is the key to respond to COVID-19 since medicine and healthcare are intertwined with data science, said experts during an online international forum hosted by the Global Strategy Institute at KAIST on April 22.
KAIST launched its think-tank, the Global Strategy Institute (GSI), in February. The GSI aims to identify global issues proactively and help make breakthroughs well aligned with solid science-based policies.
The inaugural forum of the GSI focused on how the COVID-19 pandemic would impact socio-economic, scientific, and political landscapes, under the theme “Global Cooperation in the Coronavirus Era.”
In his opening remarks, KAIST President Sung-Chul Shin stressed that future global governance will be dominated by the power of science and technology. “If we can implement efficient policies together with troubleshooting technology for responding to future crises, we will emerge stronger than before,” he said.
President Shin said ‘the Korean model’, which is being recognized as a shining example for dealing with the pandemic, is the result of collaborations combining the creativity of the private sector, the public sector’s strong infrastructure, and the full support of the citizens. He added, “Without the technological prowess coming from the competent R&D power of Korea, we could not achieve these impressive results.”
“Creative collaboration among the private and public sectors, along with research universities from around the world, will help shore up global resilience against the epidemic. We should work together to build a world of growing prosperity,” President Shin said.
Prime Minister Sye-Kyun Chung, who is in charge of the Central Disaster and Safety Countermeasures Headquarters in Korea, stressed global solidarity in his welcoming remarks, saying that “We need to share information and rely on the strength of our connections, rather than retreating into nationalistic isolation.”
Peter Lee, Vice President of the Microsoft Healthcare, pointed out in his welcoming remarks three critical sectors for global cooperation: medicine and healthcare, public health and prevention, and life and the economy. He emphasized the rule of thumb for managing data, saying that data in these fields should be open, standardized, and shared among countries to combat this global pandemic.
During a keynote session, Director General of the International Vaccine Institute (IVI) Jerome Kim described the challenges that go along with developing a vaccine. Dr. Kim said that only 7% of vaccine candidates go through the clinical trial stages, and it will take five to 10 years to completely prove a new vaccine’s safety after completing three stages of clinical tests.
“It’s very challenging to develop the vaccine for COVID-19 within 12 to 15 months,” said Dr. Kim. He added that 78 out of 115 candidates are currently undergoing clinical trials around the world.
There are five groups, including Moderna, Inovio, Jenner Institute, CanSino, and the Beijing Institute of Biological Products, who are doing clinical trials in phases 1 and 2. “Given the fact that COVID-19 is a totally new type of virus, various stakeholders’ participation, such as the National Immunization Technical Advisory Groups, the WHO, and UNICEF, is needed to work together to benefit the entire world,” he pointed out.
Professor Edward Yoonjae Choi from the Graduate School of AI at KAIST shared how AI and data sciences are being utilized to interpret the major trends of the epidemic. His group mainly focuses on deep learning to model electronic health records (EHR) for disease predictions. Professor Choi said AI and machine learning would be crucial solutions and collaborative research projects will surely accelerate how quickly we can overcome the pandemic.
In addition, Dr. Kijung Shin’s group is interpreting the SIR (Susceptible, Infected, and Recovered) model in Korea to predict the number of infections and when people were infected. However, researchers noticed that they could not see the typical modeling in Korea for predicting the number of infections since the model disregarded the new variable of humans’ efforts to stop the spread the virus.
According to research by Professor Steven Whang’s group on social distancing and face mask distribution among vulnerable age groups, people in their 20s, 60s, and 70s followed the social distancing guidelines the most strictly. The research team analyzed the data provided by SK Telecom in the Gangnam district of Seoul. The data provided on people in their 70s, a group that accounted for half of all fatalities, showed that masks were generally well distributed nationwide.
Dr. Alexandros Papaspyrids, Tertiary Education Industry Director of the Asia region of Microsoft, said that despite all the disadvantages and problems related to remote education, we shouldn’t expect to return to the days before the COVID-19 any time soon. “We should accept the new normal and explore new opportunities in the new educational environment,” he said.
Hongtaek Yong, Deputy Minister at the Office of R&D Policy at the Ministry of Science and ICT presented the Korean government’s disease prevention and response policy and how they tried to mitigate the economic and social impact.
He stressed the government’s fast testing, tracing, and openness for successfully flattening the curve, adding that the government used an ICT-based approach in all aspects of their response. From early this year when the first patient was reported, the government aggressively encouraged the biotech industry to develop diagnostic kits and novel therapeutic medications.
As a result, five companies were able to produce genetic diagnostic reagents through the emergency approval. More notably, four of them are conducting massive R&D projects sponsored by the government and this is the result of the government’s continuous investment in R&D. Korea is the leader in R&D investment among the OECD countries.
According to Yong, the government’s big data project that was launched in 2017 continuously traces the trends of epidemics in Korea. The epidemiological studies based on the paths taken by suspected patients using credit card transaction made the difference in predicting the spread of the coronavirus and implementing countermeasures. The data has been provided to the Korea’s Center for Disease Control (CDC).
“In addition to the epidemics, we have so many other pending issues arising from digital and social equities, un-contact services, and job security. We are very open to collaborate and cooperate with other countries to deal with this global crisis,” Yong said.
During the subsequent panel discussions, David Dollar, a senior fellow at the Brookings Institution, said, “The global economy in the coronavirus era will not have a rapid V-shaped recovery, but rather will fall into a long depression for at least two years.” He pointed out that if countries practice protectionism like they did during the Great Depression, the recession will be even worse. Hence, he urged the international community, especially developed nations, to avoid protectionism, consider the economic difficulties of developing countries, and provide them with financial support.
Co-Director of the Center for Universal Education at the Brookings Institution Rebecca Winthrop raised concerns over the recent shift to online teaching and learning, claiming that insufficient infrastructures in low-income families in developing nations are already causing added educational disparities and provoking the inequity issue around the world. “The ways to provide quality education equally through faster and more effective means should be studied,” she said.
Professor Joungho Kim, the director of the KAIST GSI and the forum’s organizer, concluded the event by saying that this forum will be a valuable resource for everyone who is providing assistance to those in need, both during and after the COVID-19 pandemic.
(END)
2020.04.22 View 18270 -
A Global Campaign of ‘Facts before Rumors’ on COVID-19 Launched
- A KAIST data scientist group responds to facts and rumors on COVID-19 for global awareness of the pandemic. -
Like the novel coronavirus, rumors have no borders. The world is fighting to contain the pandemic, but we also have to deal with the appalling spread of an infodemic that is as contagious as the virus. This infodemic, a pandemic of false information, is bringing chaos and extreme fear to the general public.
Professor Meeyoung Cha’s group at the School of Computing started a global campaign called ‘Facts before Rumors,’ to prevent the spread of false information from crossing borders. She explained, “We saw many rumors that had already been fact-checked long before in China and South Korea now begin to circulate in other countries, sometimes leading to detrimental results. We launched an official campaign, Facts before Rumors, to deliver COVID-19-related facts to countries where the number of cases is now increasing.” She released the first set of facts on March 26 via her Twitter account @nekozzang.
Professor Cha, a data scientist who has focused on detecting global fake news, is now part of the COVID-19 AI Task Force at the Global Strategy Institute at KAIST. She is also leading the Data Science Group at the Institute for Basic Science (IBS) as Chief Investigator.
Her research group worked in collaboration with the College of Nursing at Ewha Woman’s University to identify 15 claims about COVID-19 that circulated on social networks (SNS) and among the general public. The team fact-checked these claims based on information from the WHO and CDCs of Korea and the US. The research group is now working on translating the list of claims into Portuguese, Spanish, Persian, Chinese, Amharic, Hindi, and Vietnamese. Delivering facts before rumors, the team says, will help contain the disease and prevent any harm caused by misinformation.
The pandemic, which spread in China and South Korea before arriving in Europe and the US, is now moving into South America, Africa, and Southeast Asia. “We would like to play a part in preventing the further spread of the disease with the provision of only scientifically vetted, truthful facts,” said the team.
For this campaign, Professor Cha’s team investigated more than 200 rumored claims on COVID-19 in China during the early days of the pandemic. These claims spread in different levels: while some were only relevant locally or in larger regions of China, others propagated in Asia and are now spreading to countries that are currently most affected by the disease.
For example, the false claim which publicized that ‘Fireworks can help tame the virus in the air’ only spread in China. Other claims such as ‘Eating garlic helps people overcome the disease’ or ‘Gargling with salt water prevents the contraction of the disease,’ spread around the world even after being proved groundless.
The team noted, however, that the times at which these claims propagate are different from one country to another. “This opens up an opportunity to debunk rumors in some countries, even before they start to emerge,” said Professor Cha.
Kun-Woo Kim, a master’s candidate in the Department of Industrial Design who joined this campaign and designed the Facts before Rumors chart also expressed his hope that this campaign will help reduce the number of victims. He added, “I am very grateful to our scientists who quickly responded to the Fact Check in these challenging times.”
2020.03.27 View 10703 -
COVID-19 Map Shows How the Global Pandemic Moves
- A School of Computing team facilitated the data from COVID-19 to show the global spread of the virus. -
The COVID-19 map made by KAIST data scientists shows where and how the virus is spreading from China, reportedly the epicenter of the disease.
Professor Meeyoung Cha from the School of Computing and her group facilitated data based on the number of confirmed cases from January 22 to March 22 to analyze the trends of this global epidemic. The statistics include the number of confirmed cases, recoveries, and deaths across major continents based on the number of confirmed case data during that period.
The moving dot on the map strikingly shows how the confirmed cases are moving across the globe. According to their statistics, the centroid of the disease starts from near Wuhan in China and moved to Korea, then through the European region via Italy and Iran.
The data is collected by a graduate student from the School of Computing, Geng Sun, who started the process during the time he was quarantined since coming back from his home in China. An undergraduate colleague of Geng's, Gabriel Camilo Lima who made the map, is now working remotely from his home in Brazil since all undergraduate students were required to move out of the dormitory last week. The university closed all undergraduate housing and advised the undergraduate students to go back home in a preventive measure to stop the virus from spreading across the campus.
Gabriel said he calculated the centroid of all confirmed cases up to a given day. He explained, “I weighed each coordinate by the number of cases in that region and country and calculated an approximate center of gravity.”
“The Earth is round, so the shortest path from Asia to Europe is often through Russia. In early March, the center of gravity of new cases was moving from Asia to Europe. Therefore, the centroid is moving to the west and goes through Russia, even though Russia has not reported many cases,” he added.
Professor Cha, who is also responsible for the Data Science Group at the Institute for Basic Science (IBS) as the Chief Investigator, said their group will continue to update the map using public data at https://ds.ibs.re.kr/index.php/covid-19/.
(END)
2020.03.27 View 10293