Researchers from the Changchun Institute of Optics, Fine Mechanics and Physics of the Chinese Academy of Sciences have developed ETATrack, a single-stream network that combines visible-light and thermal-infrared images with selected historical frames. The study is published in the journal Infrared Physics & Technology.Researchers from the Changchun Institute of Optics, Fine Mechanics and Physics of the Chinese Academy of Sciences have developed ETATrack, a single-stream network that combines visible-light and thermal-infrared images with selected historical frames. The study is published in the journal Infrared Physics & Technology.Hi Tech & Innovation[#item_full_content]

Understanding speech from lip movements alone is difficult, particularly in noisy environments. A research team at the Korea Advanced Institute of Science and Technology (KAIST) has developed an AI technology that first recognizes words it can identify with high confidence and then uses the surrounding context to resolve ambiguous parts of the sentence.Understanding speech from lip movements alone is difficult, particularly in noisy environments. A research team at the Korea Advanced Institute of Science and Technology (KAIST) has developed an AI technology that first recognizes words it can identify with high confidence and then uses the surrounding context to resolve ambiguous parts of the sentence.Machine learning & AI[#item_full_content]

Nvidia CEO Jensen Huang resigned Thursday, capping a spectacular downfall for a poster boy of the artificial intelligence boom whose exit following reports of extramarital affairs now thrusts the chipmaker into an uncertain future.Nvidia CEO Jensen Huang resigned Thursday, capping a spectacular downfall for a poster boy of the artificial intelligence boom whose exit following reports of extramarital affairs now thrusts the chipmaker into an uncertain future.Business[#item_full_content]

Large language models have transformed natural language processing through their ability to generate and interpret humanlike text. Most modern language models rely on transformer architectures, whose self-attention mechanisms provide powerful capabilities for learning relationships within long sequences. However, training and running these models can require substantial computational resources and energy, creating challenges for their wider use in embedded systems, edge devices and other environments where computing resources are limited.Large language models have transformed natural language processing through their ability to generate and interpret humanlike text. Most modern language models rely on transformer architectures, whose self-attention mechanisms provide powerful capabilities for learning relationships within long sequences. However, training and running these models can require substantial computational resources and energy, creating challenges for their wider use in embedded systems, edge devices and other environments where computing resources are limited.Hardware[#item_full_content]

Researchers at the University of Tartu Institute of Computer Science are working with the Estonian technology company GScan to develop AI-based methods that use measurement data collected with the help of cosmic particles to locate objects more accurately. The technology is useful for assessing the condition of bridges and other large structures and identifying potential damage.Researchers at the University of Tartu Institute of Computer Science are working with the Estonian technology company GScan to develop AI-based methods that use measurement data collected with the help of cosmic particles to locate objects more accurately. The technology is useful for assessing the condition of bridges and other large structures and identifying potential damage.Computer Sciences[#item_full_content]

Practically or theoretically, whether teaching a classroom full of Penn State students or pondering concepts with a smaller group of international colleagues, Heather Shoenberger brings curiosity and optimism as she considers the challenges and opportunities related to artificial intelligence.Practically or theoretically, whether teaching a classroom full of Penn State students or pondering concepts with a smaller group of international colleagues, Heather Shoenberger brings curiosity and optimism as she considers the challenges and opportunities related to artificial intelligence.Machine learning & AI[#item_full_content]

A new neural network component, developed at University of Michigan Engineering, challenges a common assumption in robot learning: that the same general-purpose neural networks used for image recognition are equally suited to controlling physical movement.A new neural network component, developed at University of Michigan Engineering, challenges a common assumption in robot learning: that the same general-purpose neural networks used for image recognition are equally suited to controlling physical movement.Robotics[#item_full_content]

The control of large-scale industrial bioreactors is one of the most complex tasks in the process industry. In the often huge plants, the smallest differences in temperature, pH value or oxygen content determine the success or failure of a process.The control of large-scale industrial bioreactors is one of the most complex tasks in the process industry. In the often huge plants, the smallest differences in temperature, pH value or oxygen content determine the success or failure of a process.Engineering[#item_full_content]

A research team led by professor Zhu Hui from the Institute of Intelligent Machines at the Hefei Institutes of Physical Science of the Chinese Academy of Sciences has developed two methods to improve reconstruction fidelity, helping robots build more accurate representations of their surroundings.A research team led by professor Zhu Hui from the Institute of Intelligent Machines at the Hefei Institutes of Physical Science of the Chinese Academy of Sciences has developed two methods to improve reconstruction fidelity, helping robots build more accurate representations of their surroundings.Automotive[#item_full_content]

If “to err is human,” you could say robots are superhuman. They’re more error-prone than we are, even when tasked with things like home and factory chores, which we might consider easy. One reason for these mistakes is that robots struggle to extract relevant details from the past and act on them in real time, even with vision-language models (VLMs), which interpret visual inputs such as videos and images along with natural language.If “to err is human,” you could say robots are superhuman. They’re more error-prone than we are, even when tasked with things like home and factory chores, which we might consider easy. One reason for these mistakes is that robots struggle to extract relevant details from the past and act on them in real time, even with vision-language models (VLMs), which interpret visual inputs such as videos and images along with natural language.Robotics[#item_full_content]