RESUMO
BACKGROUND: Invasive electrophysiology (EP) training requires intellectual skills related to the interpretation of intracardiac electrograms. The classic approach to the education of young electrophysiologists focused solely on theoretical knowledge and overseen procedures in patients as no real-life-like simulation of EP studies was available. OBJECTIVE: The purpose of this study was to assess a novel tool for EP training based on fully interactive, online simulator providing real clinical experience to the users. METHODS: EP simulator users access a system with simulated electrocardiogram, mimicking signals recorded by a catheter. Assessment of EP simulator by 40 electrophysiologists from 16 countries was collected via online questionnaire. RESULTS: The realism of ECG signals was described as excellent or very good by 90% of responders, of intracardial signals by 82.5%. Realism of signal interactions and user experience was judged as excellent or very good by 75% and 70% accordingly. One hundred percent of users agree definitely or mostly that EP Simulator helps to translate theoretical into practical knowledge. Of responders, 97.5% would include it in EP training programs as it is extremely or very useful for training purposes in the opinion of 87.5%. Of responders, 72.5% think that training on EP simulator can potentially reduce the rate of complications. In 87.5%, the overall experience was completely or mostly satisfying and would be recommended by 100% of responders. CONCLUSION: EP simulator is a feasible tool for training of young electrophysiologist, and it may be potentially included in the cardiologist curriculum. We should particularly emphasize the positive respondents' assessment of EP simulator overall realism.
RESUMO
The automatic detection of violent actions in public places through video analysis is difficult because the employed Artificial Intelligence-based techniques often suffer from generalization problems. Indeed, these algorithms hinge on large quantities of annotated data and usually experience a drastic drop in performance when used in scenarios never seen during the supervised learning phase. In this paper, we introduce and publicly release the Bus Violence benchmark, the first large-scale collection of video clips for violence detection on public transport, where some actors simulated violent actions inside a moving bus in changing conditions, such as the background or light. Moreover, we conduct a performance analysis of several state-of-the-art video violence detectors pre-trained with general violence detection databases on this newly established use case. The achieved moderate performances reveal the difficulties in generalizing from these popular methods, indicating the need to have this new collection of labeled data, beneficial for specializing them in this new scenario.
Assuntos
Inteligência Artificial , Benchmarking , Violência , Algoritmos , AgressãoRESUMO
Tracking and action-recognition algorithms are currently widely used in video surveillance, monitoring urban activities and in many other areas. Their development highly relies on benchmarking scenarios, which enable reliable evaluations/improvements of their efficiencies. Presently, benchmarking methods for tracking and action-recognition algorithms rely on manual annotation of video databases, prone to human errors, limited in size and time-consuming. Here, using gained experiences, an alternative benchmarking solution is presented, which employs methods and tools obtained from the computer-game domain to create simulated video data with automatic annotations. Presented approach highly outperforms existing solutions in the size of the data and variety of annotations possible to create. With proposed system, a potential user can generate a sequence of random images involving different times of day, weather conditions, and scenes for use in tracking evaluation. In the design of the proposed tool, the concept of crowd simulation is used and developed. The system is validated by comparisons to existing methods.