SymbioLCD: Ensemble-Based Loop Closure Detection using CNN-Extracted Objects and Visual Bag-of-Words

  title={SymbioLCD: Ensemble-Based Loop Closure Detection using CNN-Extracted Objects and Visual Bag-of-Words},
  author={Jonathan J.Y. Kim and Martin Urschler and Patricia J. Riddle and J{\"o}rg Simon Wicker},
  journal={2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
Loop closure detection is an essential tool of Simultaneous Localization and Mapping (SLAM) to minimize drift in its localization. Many state-of-the-art loop closure detection (LCD) algorithms use visual Bag-of-Words (vBoW), which is robust against partial occlusions in a scene but cannot perceive the semantics or spatial relationships between feature points. CNN object extraction can address those issues, by providing semantic labels and spatial relationships between objects in a scene… 


Compressed Holistic ConvNet Representations for Detecting Loop Closures in Dynamic Environments
This paper proposes a flexible loop closure detection workflow based on the holistic ConvNet representations and introduces a compression ratio to determine how much information will be retained depending on background change and moving objects.
Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
This paper proposes a simple and scalable detection algorithm that improves mean average precision (mAP) by more than 30% relative to the previous best result on VOC 2012 -- achieving a mAP of 53.3%.
Fast and Incremental Method for Loop-Closure Detection Using Bags of Visual Words
This work presents an online method that makes it possible to detect when an image comes from an already perceived scene using local shape and color information, and extends the bag-of-words method used in image classification to incremental conditions and relies on Bayesian filtering to estimate loop-closure probability.
Semantic Mapping with Simultaneous Object Detection and Localization
This work presents a filtering-based method for semantic mapping to simultaneously detect objects and localize their 6 degree-of-freedom pose and demonstrates that the particle filtering based inference of CT-Map provides improved object detection and pose estimation with respect to baseline methods.
Graph-Based Place Recognition in Image Sequences with CNN Features
This work proposes a graph-based visual place recognition method that is able to obtain significantly better performance than that of FAB-MAP, a commonly used method for place recognition based on handcrafted features, especially on some challenging datasets.
DynaSLAM: Tracking, Mapping, and Inpainting in Dynamic Scenes
DynaSLAM is a visual SLAM system that, building on ORB-SLAM2, adds the capabilities of dynamic object detection and background inpainting, and outperforms the accuracy of standard visualSLAM baselines in highly dynamic scenarios.
RESLAM: A real-time robust edge-based SLAM system
This work builds a complete SLAM pipeline with camera pose estimation, sliding window optimization, loop closure and relocalisation capabilities that utilizes edges throughout all steps, and introduces an edge-based verification for loop closures that can also be applied for relocalisations.
On the performance of ConvNet features for place recognition
It is confirmed that networks trained for semantic place categorization also perform better at (specific) place recognition when faced with severe appearance changes and provide a reference for which networks and layers are optimal for different aspects of the place recognition problem.
Mask R-CNN
This work presents a conceptually simple, flexible, and general framework for object instance segmentation that outperforms all existing, single-model entries on every task, including the COCO 2016 challenge winners.
LSD-SLAM: Large-Scale Direct Monocular SLAM
A novel direct tracking method which operates on \(\mathfrak{sim}(3)\), thereby explicitly detecting scale-drift, and an elegant probabilistic solution to include the effect of noisy depth values into tracking are introduced.