Who write a state of art of immersive technology in education ?5 answersA state-of-the-art review of immersive technology in education was conducted by various researchers. Gampe, Vauderwange, Haiss, and Curticapean presented a conceptual immersive application for learning, focusing on augmented reality's integration with interactive simulations. Selvakumar and Sivakumar explored the potential of immersive learning technologies like virtual reality and augmented reality in revolutionizing education, emphasizing their benefits and challenges. Kuzmin and Lavlinsky discussed the application of immersive information technologies, combining augmented and virtual reality, in educational settings, highlighting the advantages of using augmented reality in textbooks. Khukalenko, Bazhina, and Zemtsov identified the expectations of the Russian pedagogical community regarding educational products based on virtual and augmented reality technologies, emphasizing factors like compliance with educational standards and high student engagement.
What is the subtopic of crowd analysis?5 answersThe subtopic of crowd analysis is the study of crowd behavior. This includes analyzing crowd statistics, such as the level of service in a crowded scene, and crowd behavior analysis, which involves studying the motion patterns and activities observed in a scene. An important aspect of crowd analysis is anomaly detection, which focuses on identifying abnormal behavior within a crowd.
What is the state of art of traffic flow (load) simulation?5 answersTraffic flow (load) simulation has been an important area of research in transportation engineering. Several limitations have been identified in existing models, including the lack of consistency among macroscopic and microscopic models, the inability to account for driver heterogeneity, the inability to look ahead into the near future, and the limitation to one-dimensional traffic. Strategies to address these limitations have been proposed, focusing on general approaches rather than specific models. One approach involves integrating machine vision with weigh-in-motion systems to improve the accuracy of traffic flow load (TFL) modeling. This approach uses deep learning methods for accurate detection of vehicles and wheels, and statistical distributions of key parameters are determined based on long-term monitoring values. An intelligent TFL model is derived from the Intelligent Driver Model (IDM), considering car-following behavior, and a simulation method is proposed to achieve accurate TFL simulation. Another approach involves modeling and simulating the movement of vehicles in established transportation infrastructures using a stochastic model based on graph and Markov chain theories. This model provides a transition probability matrix and a two-dimensional stationary distribution, which can handle the traffic's dynamic and the vehicles' distribution. The weighted least squares estimation method is applied to estimate this parameter matrix using trajectory data. Additionally, a description of driver behavior and the characteristics of macro-traffic flow have been provided, which can serve as a basis for intelligent driving and control. The simulation results show that individual adaptive driving vehicles exhibit behavior consistent with the characteristics of macro-traffic flow, demonstrating good adaptability and self-drive in changing environments and traffic conditions.
What are the challenges of using machine learning for crowd management?5 answersMachine learning for crowd management faces several challenges. One challenge is that many existing methods are problem-specific, meaning they are not easily adaptable to different crowd scenarios. This limits their effectiveness and requires additional training samples from diverse videos. Another challenge is the need for large datasets for training and testing deep learning models. Collecting diverse datasets with manual annotations and increasing video and image diversity is time-consuming. Additionally, the constrained resources of mobile devices in a mobile network, such as energy, CPU, and wireless bandwidth, pose challenges for implementing machine learning models for crowd management. The server needs to determine optimal decisions on resource management without prior knowledge of network dynamics. These challenges highlight the need for more generalized and adaptable machine learning models for crowd management.
How are crowd simulation models assessed?5 answersCrowd simulation models are assessed using various methods. One approach is to measure the similarity of the simulated crowd to the real-world crowd from a macro level using a multi-feature distribution distance algorithm. Another method involves incorporating actual crowd movement measurements and route choice models to create more accurate simulations. Additionally, a "Turing test" for crowds can be used, where human observers are presented with both real and simulated crowds and asked to distinguish between them. Evaluating the indicators of the simulation models is also important, considering factors such as crowd modeling, motion navigation, and emotion-driven crowd animation. Furthermore, crowd simulation models can be used as a modeling tool, risk assessment tool, and optimization tool to understand crowd behavior, assess risks, and optimize building design for safer crowd movement and evacuation.
What are crowd evacuation models?4 answersCrowd evacuation models are integrated models that simulate the interaction between crowds and fluid particles during evacuation scenarios. These models treat both the crowd and the water as fluid particles, allowing for the incorporation of various forces such as pressure, shear, buoyancy, and active forces to drive the agents. They also involve the development of practical evacuation strategies by observing and studying survival techniques from whirlpools and sudden changes in water levels during floods. Additionally, these models consider factors such as panic behaviors, panic evolution, and the stress responses of pedestrians with different personality traits to panic emotion. They also analyze the influence of factors like relatives moving together or looking for each other on the chaos degree of evacuating crowds. The effectiveness of these models has been demonstrated through extensive simulation results, providing useful insights into the design of practical evacuation strategies.