A framework for converting speech into semantic navigationoperational pathIn order to make indoor service robots understand human languagebetter and navigate more accurately in indoor environment, a newframework is proposed to transform speech input into robot operationpath. Firstly, the combination of n-gram and Chinese pinyin is used todetect and correct the Chinese characters converted into speech input toensure the accuracy of the robot to execute the command. Next, we builta vocabulary about the interior to improve the effect of word segmentation.On the result of Chinese word segmentation, the noun is extractedand matched with the information node on the semantic map. Finally,according to the order of natural language description, each node isconnected by the verbs from the sentence, which forms the sequence ofoperations to be performed by the robot, which provides the basis for thesubsequent path planning on the semantic map. The framework enhancesthe robot’s ability to understand user intent and is of great significancefor human-robot interactive semantic navigation research.
