Huawei is making a bold shift from the Mate 10 to enterprise intelligence, stepping onto a new track in artificial intelligence. Recently, Google unveiled the AVA (Atomic Visual Actions) database, a collection of YouTube videos annotated with 80 atomic actions like walking, shaking hands, and kicking. Each action comes with spatial and temporal positioning, totaling 57,600 video clips, 96,000 human actions, and 210,000 action tags. Google claims this technology will not only enhance its video analysis capabilities but also help advertisers better target consumers. The ultimate goal is to teach computers social intelligence—understanding what humans are doing, what they might do next, and their intentions.
Currently, many domestic research institutions and companies are investing heavily in behavior recognition to tackle the challenges of understanding video surveillance data. However, achieving the goal of machines truly "understanding" human actions requires significant effort. According to Professor Li Jinping from Jinan University’s Pattern Recognition and Intelligent Systems Institute, most current human behavior recognition systems are still at the motion recognition stage, and recognizing human actions remains a major challenge for practical applications. He noted that while it's possible for machines to detect movement, it's still a "may have" scenario at this stage.
In terms of application, behavior recognition has shown great potential in security and other fields. For instance, in elderly care homes, a system can alert staff if an elderly person falls, even contacting hospitals directly. This technology is often likened to the "Mission Impossible" black tech. Gait recognition, which identifies individuals based on their walking style, is one of the most promising applications. Despite technical challenges, behavior recognition is already being used in various areas such as public security, energy, banking, and healthcare.
Professor Wu Fei from Zhejiang University's Institute of Artificial Intelligence explained that this technology was first applied by Apple and Microsoft in gaming and is now expanding into law enforcement, power generation, grid safety, banking, and hospital monitoring. Gait recognition, though a form of identity recognition, is a powerful intersection of behavior and identity. It involves detecting, segmenting, identifying, and tracking human movement, with the most challenging part relying heavily on behavior recognition research.
In recent years, gait recognition has played a key role in missing persons searches and identifying suspects. In the "Wisdom" program, galactic drop technology successfully identified a suspect on the spot using gait recognition. Later, it even recognized a dog's silhouette through gait. Huang Yongbiao, founder of Galaxy Drop Technology, emphasized that gait recognition excels in analyzing abnormal behaviors, which is a major advantage over other biometric features. This includes detecting body shape and movement to identify unusual activities, involving multiple technical steps such as detection, segmentation, key point location, identification, and tracking.
The challenges in behavior recognition include locating and tracking human motion in continuous video streams. Wu Fei pointed out that identifying the start and end points of actions in video is extremely difficult due to factors like background changes and camera shake. Scene recognition is another big hurdle, as different people perform the same action differently depending on time, speed, posture, and occlusion. Li Jinping added that unlike static objects, actions evolve over time, introducing more uncertainty. A single image of someone running might actually be a jumping action, making it hard to determine the exact event without multiple frames.
Google’s AVA dataset offers fine-grained temporal and spatial annotations, enabling deeper research into models and improving accuracy in human activity recognition. Looking ahead, both equipment platforms and theoretical models are gaining momentum. Li Jinping highlighted China's strong potential in behavior recognition over the next five to ten years, suggesting that supporting domestic equipment and platforms could help reduce reliance on foreign technologies. Wu Fei added that Chinese teams should focus on developing theoretical models and data-driven basic theories, emphasizing the need for better understanding of motion states and joint importance in action discrimination. These advancements will be crucial for the future of behavior recognition.
Football Stadium Advertising LED Screen
This series is a classic football field advertising LED Screen, which is specially used for advertising display in various stadiums, such as football fields, basketball courts, badminton courts, tennis courts, volleyball courts and other competitive sports halls. Football Stadium Advertising LED Screen with IP65 waterproof level, excellent anti-collision ability, can resist the impact of balls. High brightnesee makes good perfomance even under sunshine. Good protection design on top of cabinet with thick anti-collision cotton.
Football Stadium Advertising Led Screen,Led Panel Display,Led Video Panel, Football Cup Led Display Billboards
Guangzhou Cheng Wen Photoelectric Technology Co., Ltd. , https://www.cwdisplay.com