Monocular VO Based on Deep Siamese Convolutional Neural Network

Figure 1

Architecture of the end-to-end DSCNN-based monocular VO.