Robust visual SLAM across seasons

By Tayyab Naseer, Michael Ruhnke, Cyrill Stachniss, Luciano Spinello, Wolfram Burgard. 2015

103 citations recorded by the source corpus.

In this paper, we present an appearance-based visual SLAM approach that focuses on detecting loop closures across seasons. Given two image sequences, our method first extracts one descriptor per image for both sequences using a deep convolutional neural network. Then, we compute a similarity matrix by comparing each image of a query sequence with a database. Finally, based on the similarity matrix, we formulate a flow network problem and compute matching hypotheses between sequences. In this way, our approach can handle partially matching routes, loops in the trajectory and different speeds of the robot. With a matching hypothesis as loop closure information and the odometry information of the robot, we formulate a graph based SLAM problem and compute a joint maximum likelihood trajectory.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

1. Lihua Xie
2. Han Wang
3. Chen Wang
4. Lei Tai

In Computer Vision and Pattern Recognition

1. Chen Wang
2. Lihua Xie
3. Han Wang
4. Lei Tai
5. Ming Liu

Citations in the corpus

• A Survey of Deep Network Solutions for Learning Control in Robotics: From Reinforcement to Imitation (2016, neutral): “atures for visual place recognition in robotics was proved as well. It also proposed that mid and high-level features from the CNN models were more robust for the changes of appearance and viewpoint. Naseer et al. (2015) and Arroyo et al. (2016) showed that features extracted by convolutional models”
• Deep-learning in Mobile Robotics - from Perception to Control Systems: A Survey on Why and Why not (2016, background): “For reinforcement learning, it is more like an optimization problem.”
• Simultaneous Localization and Mapping: A Survey of Current Trends in Autonomous Driving (2017, unknown): “More generally, CNNs could change how place recognition is performed by using feature maps coming from these networks as in [142].”
• Intensity Scan Context: Coding Intensity and Geometry Relations for Loop Closure Detection (2020, background): “Recently some works target to solve environmental changes such as [13], [14], but are limited to some specific scenarios.”
• ReFusion: 3D Reconstruction in Dynamic Environments for RGB-D Cameras Exploiting Residuals (2019, unknown): “Other approaches exploit image sequences [9], [20].”