Without bells and whistles, VisTR achieves the highest speed among all existing VIS models, and achieves the best result among methods using single model on the YouTube-VIS dataset. For the first time, we demonstrate a much simpler and faster video instance segmentation framework built upon Transformers, achieving competitive accuracy. We hope that VisTR can motivate future research for more video understanding tasks.
Abstract-Electromagnetic wave of electrical spark is a potential cause to eletrical equipment failure. This research focused on identificating and comparative analyzing the different types of electromagnetic waveform generated by eletrical equipment failure based on SVM. After analyzing and extracting the features the electromagnetic waveform, a model was built to identificate the type of the elctromagnetic waveform. The collected standard electromagnetic waveforms were used as the imput of the train model and the model accuracy was improved by adjusting training parameters afer analyzing the results, When inputting an unknown type of electromagnetic waveform, SVM may predict the output of the network according to the recognition rule. Then the types of electromagnetic waveforms were identificated by using adjusted models. The result shows that the electromagnetic waveform can be effectively and feasibly identificated based on SVM, which provides a theoretical support on prediction method of gas explosion caused by electrical sparks.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.