Learnable Line Segment Descriptor for Visual SLAM

Abstract

Traditionally, the indirect visual motion estimation and simultaneous localization and mapping (SLAM) systems were based on point features. In recent years, several SLAM systems that use lines as primitives were suggested. Despite the extra robustness and accuracy brought by the line segment matching, the line segment descriptors used in such systems were hand-crafted, and therefore sub-optimal. In this paper, we suggest applying descriptor learning to construct line segment descriptors optimized for matching tasks. We show how such descriptors can be constructed on top of a deep yet lightweight fully-convolutional neural network. The coefficients of this network are trained using an automatically collected dataset of matching and non-matching line segments. The use of the fully-convolutional network ensures that the bulk of the computations needed to compute descriptors is shared among the multiple line segments in the same image, enabling efficient implementation. We show that the learned line segment descriptors outperform the previously suggested hand-crafted line segment descriptors both in isolation (i.e., for the subtask of distinguishing matching and non-matching line segments), but also when built into the SLAM system. We construct a new line based SLAM pipeline built upon a state-of-the-art point-only system. We demonstrate generalization of the learned parameters of the descriptor network between two well-known datasets for autonomous driving and indoor micro aerial vehicle navigation.

Hessian with Mini-Batches for Electrical Demand Prediction

“…It should be noted that the neural network trained by steepest descent with mini-batches (SDMB) in [9][10][11][12] and using (19), (22), (23) had l = 6 neurons in its hidden layer, mini-batches with a size of y = 32, a tuning factor of α = 0.0004, and a number of epochs of e = 40.…”

“…In this section, we compare steepest descent (SD), steepest descent with mini-batches (SDMB) from [9][10][11][12], the Hessian (H) from [17][18][19][20], and the Hessian with mini-batches (HMB) from this investigation for electrical demand prediction. The goal of these algorithms is that the neural network output q l must reach the target t l as soon as possible.…”

“…There are several applications for mini-batches. In [9][10][11][12], mini-batches were employed for tuning. In [13,14], mini-batches were used for clustering.…”

A Clutter-Resistant SLAM Algorithm for Autonomous Guided Vehicles in Dynamic Industrial Environment

“…Line features are widely employed for SLAM in many indoor scenes. Vakhitov et al propose a learning method based on a fully-convolutional network to construct line segment descriptors [35]. However, in some dynamic industrial environments, line features may be unavailable in a short period of time when being occluded by some movable objects, e.g.…”

FPGA Design of Real-Time MDFD System Using High Level Synthesis

“…Image 3D information extraction is an important part of computer vision systems. In the increasingly mature artificial intelligence system, the 2D information of the scene can no longer meet the needs of researchers, especially in the related fields of robotic arm control [1], SLAM navigation technology [2] and other fields [3] in need of three-dimensional information.…”