Hunting for Naval Mines with Deep Neural Networks
OCEANS-IEEE(2017)
摘要
Explosive naval mines pose a threat to ocean and sea faring vessels, both military and civilian. This work applies deep neural network (DNN) methods to the problem of detecting mine-like objects (MLO) on the seafloor in side-scan sonar imagery. We explored how the DNN depth, memory requirements, calculation requirements, and training data distribution affect detection efficacy. A visualization technique (class activation map) was incorporated that aids a user in interpreting the model's behavior. We found that modest DNN model sizes yielded better accuracy (98%) than very simple DNN models (93%) and a support vector machine (78%). The largest DNN models achieved <1% efficacy increase at a cost of a 17x increase of trainable parameter count and computation requirements. In contrast to DNNs popularized for many-class image recognition tasks, the models for this task require far fewer computational resources (0.3% of parameters), and are suitable for embedded use within an autonomous unmanned underwater vehicle.
更多查看译文
关键词
DNN models,image recognition tasks,data distribution training,autonomous unmanned underwater vehicle,modest DNN model,visualization technique,detection efficacy,calculation requirements,memory requirements,DNN depth,side-scan sonar imagery,minelike objects,deep neural network methods,sea faring vessels,explosive naval mines,deep neural networks,computation requirements,trainable parameter count,support vector machine
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要