Abstract
Achieving high estimation accuracy is significant for semantic simultaneous localization and mapping (SLAM) tasks. Yet, the estimation process is vulnerable to several sources of error, including limitations of the instruments used to perceive the environment, shortcomings of the employed algorithm, environmental conditions, or other unpredictable noise. In this article, a novel stacked long short-term memory (LSTM)-based error reduction approach is developed to enhance the accuracy of semantic SLAM in presence of such error sources. Training and testing data sets were constructed through simulated and real-time experiments. The effectiveness of the proposed approach was demonstrated by its ability to capture and reduce semantic SLAM estimation errors in training and testing data sets. Quantitative performance measurement was carried out using the absolute trajectory error (ATE) metric. The proposed approach was compared with vanilla and bidirectional LSTM networks, shallow and deep neural networks, and support vector machines. The proposed approach outperforms all other structures and was able to significantly improve the accuracy of semantic SLAM. To further verify the applicability of the proposed approach, it was tested on real-time sequences from the TUM RGB-D data set, where it was able to improve the estimated trajectories.
Highlights
S IMULTANEOUS localization and mapping (SLAM) is one of the most prevalent research problems in the robotics community
Deep learning-based object detection techniques [25]–[27] promoted the advancement of object-based semantic simultaneous localization and mapping (SLAM), which relies on observations of landmarks that can be semantically labeled in the environment, such as the approaches presented in [28] and [29]
The results have proven the ability of the proposed approach to successfully identify and reduce pose estimation errors resulting from multiple factors in the semantic SLAM pipeline
Summary
Rana Azzam , Yusra Alkendi , Tarek Taha , Shoudong Huang , Senior Member, IEEE, and Yahya Zweiri , Member, IEEE. Abstract— Achieving high estimation accuracy is significant for semantic simultaneous localization and mapping (SLAM) tasks. A novel stacked long short-term memory (LSTM)-based error reduction approach is developed to enhance the accuracy of semantic SLAM in presence of such error sources. Training and testing data sets were constructed through simulated and real-time experiments. The effectiveness of the proposed approach was demonstrated by its ability to capture and reduce semantic SLAM estimation errors in training and testing data sets. The proposed approach outperforms all other structures and was able to significantly improve the accuracy of semantic SLAM. To further verify the applicability of the proposed approach, it was tested on real-time sequences from the TUM RGB-D data set, where it was able to improve the estimated trajectories
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have