首页>
外国专利>
LEARNING METHOD AND LEARNING DEVICE FOR UPDATING HD MAP BY RECONSTRUCTING 3D SPACE BY USING DEPTH ESTIMATION INFORMATION AND CLASS INFORMATION ON EACH OBJECT, WHICH HAVE BEEN ACQUIRED THROUGH V2X INFORMATION INTEGRATION TECHNIQUE, AND TESTING METHOD AND TESTING DEVICE USING THE SAME
LEARNING METHOD AND LEARNING DEVICE FOR UPDATING HD MAP BY RECONSTRUCTING 3D SPACE BY USING DEPTH ESTIMATION INFORMATION AND CLASS INFORMATION ON EACH OBJECT, WHICH HAVE BEEN ACQUIRED THROUGH V2X INFORMATION INTEGRATION TECHNIQUE, AND TESTING METHOD AND TESTING DEVICE USING THE SAME
A learning method for selecting specific information to be used to update an HD map, comprising: (a) a learning device, causing a coordinate neural network to generate a local feature map and a global feature vector by applying a coordinate neural network operation to a coordinate matrix step; (b) causing, by the learning apparatus, a decision neural network to generate a first prediction fitness score to an N-th prediction fitness score by applying a decision neural network operation to the integrated feature map; and (c) the learning device causes the loss layer to obtain (i) the first prediction fitness score to the N-th prediction fitness score and (ii) the first ground-truth (GT) fitness score to the N-th fitness score. A method comprising; generating a loss with reference and learning the parameters of the decision neural network and the coordinate neural network by performing backpropagation using the loss is provided.
展开▼