Scientists adopt deep learning for multi-object tracking
Research News Release
EurekAlert! provides eligible reporters with free access to embargoed and breaking news releases.Eligibility Guidelines
EurekAlert! offers eligible public information officers paid access to a reliable news release distribution service.Eligibility Guidelines
EurekAlert! is a service of the American Association for the Advancement of Science.
Implementing algorithms that can simultaneously track multiple objects is essential to unlock many applications, from autonomous driving to advanced public surveillance. However, it is difficult for computers to discriminate between detected objects based on their appearance. Now, researchers at the Gwangju Institute of Science and Technology (GIST) adapted deep learning techniques in a multi-object tracking framework, overcoming short-term occlusion and achieving remarkable performance without sacrificing computational speed.
Creating new procedures that improve mass drone traffic is the purpose of LABYRINTH, a European research project coordinated by the Universidad Carlos III de Madrid (UC3M) with the participation of 13 international organisations within the R&D&I, transport, emergency, and auxiliary services fields. Researchers hope to use these drone swarm applications to improve civil road, train, sea, and air transport, making it safer, more efficient, and more sustainable.
A team of researchers, led by Yale-NUS College Associate Professor of Science (Computer Science) Robby Tan, who is also from the National University of Singapore's Faculty of Engineering, has developed novel approaches using computer vision and deep learning to resolve the problem of low-level vision in videos caused by rain and night-time conditions, as well as improve the accuracy of 3D human pose estimation in videos.
Researchers from The University of Tokyo Institute of Industrial Science have developed a machine learning-based model to predict the characteristics of bonded systems. Using the density of states of the individual component reactants, they have achieved accurate predictions of the binding energy, bond length, number of covalent electrons, and Fermi energy. The broadly applicable model is expected to make a significant contribution to the development of materials such as catalysts and nanowires.
UC Riverside engineers made a pneumatic RAM chip using microfluidic valves instead of electronic transistors. The valves remain sealed against a pressure differential even when disconnected from an air supply line, creating trapped pressure differentials that function as memories and maintain the states of a robot's actuators. Dense arrays of these valves can perform advanced operations and reduce the expensive, bulky, and power-consuming electronic hardware typically used to control pneumatic robots.
Despite advances in deep neural networks, computers still struggle with the very human skill of "imagination." Now, a USC research team has developed an AI that uses human-like capabilities to imagine a never-before-seen object with different attributes.
Vaccine negativity and reluctance is not a recent phenomenon but, to date, little research has been done to explore the dominance of negative vaccine-related information.
A team of researchers from the University of Maryland has 3D printed a soft robotic hand that is agile enough to play Nintendo's Super Mario Bros. - and win!
Scientists have invented a first-of-its-kind instrument to peer deeply into billions of Twitter posts--providing an unprecedented, minute-by-minute view of popularity, from rising political movements, to K-pop, to emerging diseases. The tool--called the Storywrangler--gathers phrases across 150 different languages, analyzing the rise and fall of ideas and stories, each day, among people around the world. The Storywrangler quantifies collective attention.
Northwestern University engineers have developed the first full, three-dimensional (3D), dynamic simulation of a rat's complete whisker system, offering rare, realistic insight into how rats obtain tactile information.