Lightweight Unsupervised Deep Loop Closure

Abstract

Robust efficient loop closure detection is essential for large-scale real-time SLAM. In this paper, we propose a novel unsupervised deep neural network architecture of a feature embedding for visual loop closure that is both reliable and compact. Our model is built upon the autoencoder architecture, tailored specifically to the problem at hand. To train our network, we inflict random noise on our input data as the denoising autoencoder does, but, instead of applying random dropout, we warp images with randomized projective transformations to emulate natural viewpoint changes due to robot motion. Moreover, we utilize the geometric information and illumination invariance provided by histogram of oriented gradients (HOG), forcing the encoder to reconstruct a HOG descriptor instead of the original image. As a result, our trained model extracts features robust to extreme variations in appearance directly from raw images, without the need for labeled training data or environment-specific training. We perform extensive experiments on various challenging datasets, showing that the proposed deep loop-closure model consistently outperforms the state-of-the-art methods in terms of effectiveness and efficiency. Our model is fast and reliable enough to close loops in real time with no dimensionality reduction, and capable of replacing generic off-the-shelf networks in state-of-the-art ConvNet-based loop closure systems.

arXiv commentPublished in Proc. of Robotics: Science and Systems 2018

References in corpus (5)

• Caffe: Convolutional Architecture for Fast Feature Embedding
• Experiments with Random Projection
• Convolutional Neural Network-based Place Recognition
• On the Performance of ConvNet Features for Place Recognition
• Deep Learning Features at Scale for Visual Place Recognition

Cited by in corpus (7)

• A Survey of Simultaneous Localization and Mapping with an Envision in 6G Wireless Networks
• Are State-of-the-art Visual Place Recognition Techniques any Good for Aerial Robotics?
• Learning Whole-Image Descriptors for Real-time Loop Detection andKidnap Recovery under Large Viewpoint Difference
• Sequence-Based Filtering for Visual Route-Based Navigation: Analysing the Benefits, Trade-offs and Design Choices
• An Efficient Schmidt-EKF for 3D Visual-Inertial SLAM
• A Benchmark Comparison of Visual Place Recognition Techniques for Resource-Constrained Embedded Platforms
• Place recognition survey: An update on deep learning approaches