Author: Poosarla Jayanth and Rajeev Yadav
Download PDF
Drones can deliver payloads, acquire real-time data in an efficient and cost-effective manner, and have been a driving force behind the rapid development of a wide variety of industrial, commercial, and recreational applications. In order to attend the requirements of living human people in terms of medical services or support, drones play a crucial role. The study met the challenges related to model's weights that iteratively changed depending on the computed loss and gradients with respect to the loss. The optimizer makes adjustments to the weights in order to reduce the amount of loss and improve the model's overall performance on the given task. Regrettably, advancements in medical science have occurred at a more glacial pace in recent years. The primary objective of this research is to carry out a study on the application of AI-based robotic drones to emergency medical situations. This kind of visualisation enables you to keep track of how well your model is picking up new information during training and to see any problems like over- or under fitting. The accuracy rate was found to be 0.8530% and the loss function was found to be 0.4625% based on the research's findings.
Artificial Intelligence, Drones, Medical Applications, Robotics
A U-Net model built with Tensor Flow's Keras application programming interface (API). The configuration of the model's training procedure is accomplished through the usage of the compile function. After the model has been compiled, it is now prepared to be trained with the help of the fit function. During training, the model's weights are iteratively changed depending on the computed loss and gradients with respect to the loss. This process takes place in the background. The optimizer makes adjustments to the weights in order to reduce the amount of loss and improve the model's overall performance on the given task. During training, the metrics that are given in the compile function are calculated and shown. This is done in order to provide insights into the performance of the model as it learns from the training data. According to the findings of this investigation, the accuracy rate was found to be 0.8530%, and the loss function was found to be 0.4625%.
-
-
Indexing