Monocular Simultaneous Localization And Mapping System Research Articles

LDVI-SLAM: A Lightweight Monocular Visual-Inertial SLAM System for Dynamic Environments Based on Motion Constraints

Abstract Traditional Simultaneous Localization and Mapping (SLAM) systems are typically based on the assumption of a static environment. However, in practical applications, the presence of moving objects significantly reduces localization accuracy, limiting the system's versatility. To address the challenges of SLAM systems in dynamic environments, the academic community often employs computationally intensive methods such as deep learning, and some algorithms rely on expensive sensors (e.g., LiDAR or RGBD cameras) to obtain depth information. These factors increase computational complexity or hardware costs, complicating practical deployment. To improve localization accuracy and adaptability of SLAM systems in dynamic scenarios while maintaining low deployment costs, this paper proposes a dynamic environment robust monocular inertial SLAM system named LDVI-SLAM. The system uses more cost-effective sensors—monocular cameras and Inertial Measurement Unit (IMU)—along with lightweight computational methods. In LDVI-SLAM, first, the reliability of IMU data is verified. Then, using the ego-motion information provided by the IMU, along with epipolar constraint and an improved Rotation-aware Flow Vector Bound (R-FVB) constraint, dynamic feature points are eliminated. Additionally, this paper proposes a continuous tracking across interval frames method to enhance the distinction between static and dynamic feature points. Experimental results demonstrate that LDVI-SLAM performs effectively in dynamic environments and is easy to deploy. On the VIODE dataset, experimental results show that compared to the deep learning-based DynaSLAM, this method reduces the RMSE (Root Mean Square Error) of absolute trajectory error by 10.3%. Moreover, in terms of speed, under the same computing power, the single-frame processing speed of this method is comparable to ORB-SLAM3 and is two orders of magnitude faster than DynaSLAM, significantly outperforming deep learning-based SLAM algorithms. Experiments on the OMD dataset further prove that this method effectively avoids the risk of semantic classification errors, demonstrating better robustness and generality.

CTO-SLAM: Contour Tracking for Object-Level Robust 4D SLAM

The demand for 4D ( 3D+time ) SLAM system is increasingly urgent, especially for decision-making and scene understanding. However, most of the existing simultaneous localization and mapping ( SLAM ) systems primarily assume static environments. They fail to represent dynamic scenarios due to the challenge of establishing robust long-term spatiotemporal associations in dynamic object tracking. We address this limitation and propose CTO-SLAM, a monocular and RGB-D object-level 4D SLAM system to track moving objects and estimate their motion simultaneously. In this paper, we propose contour tracking, which introduces contour features to enhance the keypoint representation of dynamic objects and coupled with pixel tracking to achieve long-term robust object tracking. Based on contour tracking, we propose a novel sampling-based object pose initialization algorithm and the following adapted bundle adjustment ( BA ) optimization algorithm to estimate dynamic object poses with high accuracy. The CTO-SLAM system is verified on both KITTI and VKITTI datasets. The experimental results demonstrate that our system effectively addresses cumulative errors in long-term spatiotemporal association and hence obtains substantial improvements over the state-of-the-art systems. The source code is available at https://github.com/realXiaohan/CTO-SLAM.

Improving autonomous detection in dynamic environments with robust monocular thermal SLAM system

Thermal SLAM outperforms visual SLAM under conditions of nighttime, low visibility (e.g., fog, smoke, and dust), and direct glare. However, the unique non-uniformity correction for thermal imaging will cause data interruption, and the weak texture of the thermal image itself will lead to poor localization accuracy. To overcome these limitations, this paper presents a monocular thermal camera-based simultaneous localization and mapping (SLAM) system that can be used for high-precision and robust localization in challenging dynamic environments with visual degradation. We also conducted extensive experiments on small-scale indoor and outdoor sequences and on large-scale driving sequences, totaling over 180,000 real-world thermal images. Our results show that the developed MonoThermal-SLAM system can achieve camera localization and sparse structured map reconstruction in the face of visual degradation and moving objects in dynamic environments without the need for other sensors. Compared with state-of-the-art monocular SLAM systems, MonoThermal-SLAM was the only system that successfully tracked all sequences to offer unparalleled robustness and higher positioning accuracy. Relative position error (RPE) was less than 0.1 m of the ground-truth trajectory in both the indoor and outdoor sequences. Absolute trajectory error (ATE) improves averagely over 68.57% compared to other best-performing SLAM systems in driving sequences. The average relative error of tracking path length to the ground-truth was less than 5.19%. The superior performance of MonoThermal-SLAM is due to the introduction of several key innovations. First, the developed real-time scene-based denoising chain not only avoids thermal camera data interruptions but also significantly improves the thermal image quality and widens the boundaries of the application of visual SLAM systems for thermal images. Second, the combination of epipolar constraints and semantic segmentation reduces the interference of dynamic objects and improves the robustness of thermal SLAM in dynamic scenes. Third, the implemented SLAM system based on point and line features consists of well-designed initialization, tracking, local mapping, and loop closing to overcome the drawbacks of poor spatial texture distribution of thermal images. Therefore, the system can be used as a novel positioning solution to replace expensive commercial navigation systems, especially in challenging urban dynamic environments with changing light and low visibility in the air.

SVR-Net: A Sparse Voxelized Recurrent Network for Robust Monocular SLAM with Direct TSDF Mapping.

Simultaneous localization and mapping (SLAM) plays a fundamental role in downstream tasks including navigation and planning. However, monocular visual SLAM faces challenges in robust pose estimation and map construction. This study proposes a monocular SLAM system based on a sparse voxelized recurrent network, SVR-Net. It extracts voxel features from a pair of frames for correlation and recursively matches them to estimate pose and dense map. The sparse voxelized structure is designed to reduce memory occupation of voxel features. Meanwhile, gated recurrent units are incorporated to iteratively search for optimal matches on correlation maps, thereby enhancing the robustness of the system. Additionally, Gauss-Newton updates are embedded in iterations to impose geometrical constraints, which ensure accurate pose estimation. After end-to-end training on ScanNet, SVR-Net is evaluated on TUM-RGBD and successfully estimates poses on all nine scenes, while traditional ORB-SLAM fails on most of them. Furthermore, absolute trajectory error (ATE) results demonstrate that the tracking accuracy is comparable to that of DeepV2D. Unlike most previous monocular SLAM systems, SVR-Net directly estimates dense TSDF maps suitable for downstream tasks with high efficiency of data exploitation. This study contributes to the development of robust monocular visual SLAM systems and direct TSDF mapping.

Improved Real-Time Monocular SLAM Using Semantic Segmentation on Selective Frames

Monocular simultaneous localization and mapping (SLAM) is emerging in advanced driver assistance systems and autonomous driving, because a single camera is cheap and easy to install. Conventional monocular SLAM has two major challenges leading inaccurate localization and mapping. First, it is challenging to estimate scales in localization and mapping. Second, conventional monocular SLAM uses inappropriate mapping factors such as dynamic objects and low-parallax areas in mapping. This paper proposes an improved real-time monocular SLAM that resolves the aforementioned challenges by efficiently using deep learning-based semantic segmentation. To achieve the real-time execution of the proposed method, we apply semantic segmentation only to downsampled keyframes in parallel with mapping processes. In addition, the proposed method corrects scales of camera poses and three-dimensional (3D) points, using estimated ground plane from road-labeled 3D points and the real camera height. The proposed method also removes inappropriate corner features labeled as moving objects and low parallax areas. Experiments with eight video sequences demonstrate that the proposed monocular SLAM system achieves significantly improved and comparable trajectory tracking accuracy, compared to existing state-of-the-art monocular and stereo SLAM systems, respectively. The proposed system can achieve real-time tracking on a standard CPU potentially with a standard GPU support, whereas existing segmentation-aided monocular SLAM does not.

HFNet-SLAM: An Accurate and Real-Time Monocular SLAM System with Deep Features.

Image tracking and retrieval strategies are of vital importance in visual Simultaneous Localization and Mapping (SLAM) systems. For most state-of-the-art systems, hand-crafted features and bag-of-words (BoW) algorithms are the common solutions. Recent research reports the vulnerability of these traditional algorithms in complex environments. To replace these methods, this work proposes HFNet-SLAM, an accurate and real-time monocular SLAM system built on the ORB-SLAM3 framework incorporated with deep convolutional neural networks (CNNs). This work provides a pipeline of feature extraction, keypoint matching, and loop detection fully based on features from CNNs. The performance of this system has been validated on public datasets against other state-of-the-art algorithms. The results reveal that the HFNet-SLAM achieves the lowest errors among systems available in the literature. Notably, the HFNet-SLAM obtains an average accuracy of 2.8 cm in EuRoC dataset in pure visual configuration. Besides, it doubles the accuracy in medium and large environments in TUM-VI dataset compared with ORB-SLAM3. Furthermore, with the optimisation of TensorRT technology, the entire system can run in real-time at 50 FPS.

Semi-Direct Multimap SLAM System for Real-Time Sparse 3-D Map Reconstruction

The lack of a sufficient number of reliable corners in low-textured environments is a big challenge for classical visual simultaneous localization and mapping (SLAM), especially for point feature-based methods. Many other features (i.e., line and plane segments) are often combined with points to restore an environmental structure. However, using such features requires much computational time. This work focuses on the reliable and high-performance real-time operation of a SLAM system in low-textured scenarios. It proposes a semi-direct multi-map monocular SLAM system (SM-SLAM) that combines direct tracking and feature-based map maintenance with point features and line segments. The proposed system tracks non-keyframes based on a sparse image alignment method for fast tracking, extracts and matches point features and line segments in keyframes for high-quality environment structure and motion optimization. We present an extensive evaluation on two widely-used public datasets and some challenging real-world scenarios. Experimental results show that SM-SLAM can well reconstruct a sparse 3D map with geometrical structure information in 30-40 Hz. Notably, SM-SLAM shows an accuracy improvement of more than 20% than the Oriented FAST and Rotated BRIEF feature-based SLAM (ORB-SLAM) on some low-speed, small-range camera motion datasets and performs well in low-texture scenarios.

An ORB-SLAM3-based Approach for Surgical Navigation in Ureteroscopy

ABSTRACT In ureteroscopy a flexible ureteroscope is used to inspect the kidneys and the ureters. In this context, we investigate robotic solutions that leverage recent developments from Simultaneous Localization and Mapping (SLAM) to provide a 3D map of the organ to be explored and the ureteroscope’s pose. With this aid, the surgeoncan navigate through the organ more precisely. Additionally, for the final organ inspection, the risk of missing certain regions of the organ is minimized. In this paper, we propose a visual, monocular SLAM system based on ORB-SLAM3 that is able to estimate the pose of the ureteroscope’s tip. In order to fulfill this task, we introduce two pre-processing steps. The first one aims at increasing the contrast of the image and the second one helps to avoid detecting features in non-desired regions (e.g. reflections). Additionally, we extend ORB-SLAM3 with A-KAZE and SuperPoint features and compare their performances to ORB features. The proposed method is evaluated in two experiments. The first experiment shows that we are able to estimate the trajectory of the ureteroscope and the map of a synthetic kidney with low errors. The second experiment shows that the method also gives promising results in cystoscopy.

A monocular SLAM system based on SIFT features for gastroscope tracking

During flexible gastroscopy, physicians have extreme difficulties to self-localize. Camera tracking method such as simultaneous localization and mapping (SLAM) has become a research hotspot in recent years, allowing tracking of the endoscope. However, most of the existing solutions have focused on tasks in which sufficient texture information is available, such as laparoscope tracking, and cannot be applied to gastroscope tracking since gastroscopic images have fewer textures than laparoscopic images. This paper proposes a new monocular SLAM framework based on scale-invariant feature transform (SIFT) and narrow-band imaging (NBI), which extracts SIFT features instead of oriented features from accelerated segment test (FAST) and rotated binary robust independent elementary features (BRIEF) features from gastroscopic NBI images, and performs feature retention based on the response sorting strategy for achieving more matches. Experimental results show that the root mean squared error of the proposed algorithm can reach a minimum of 2.074mm, and the pose accuracy can be improved by up to 25.73% compared with oriented FAST and rotated BRIEF (ORB)-SLAM. SIFT features and response sorting strategy can achieve more accurate matching in gastroscopic NBI images than other features and homogenization strategy, and the proposed algorithm can also run successfully on real clinical gastroscopic data. The proposed algorithm has the potential clinical value to assist physicians in locating the gastroscope during gastroscopy.

A monocular visual SLAM system augmented by lightweight deep local feature extractor using in-house and low-cost LIDAR-camera integrated device

ABSTRACT Simultaneous Localization and Mapping (SLAM) has been widely used in emergency response, self-driving and city-scale 3D mapping and navigation. Recent deep-learning based feature point extractors have demonstrated superior performance in dealing with the complex environmental challenges (e.g. extreme lighting) while the traditional extractors are struggling. In this paper, we have successfully improved the robustness and accuracy of a monocular visual SLAM system under various complex scenes by adding a deep learning based visual localization thread as an augmentation to the visual SLAM framework. In this thread, our feature extractor with an efficient lightweight deep neural network is used for absolute pose and scale estimation in real time using the highly accurate georeferenced prior map database at 20cm geometric accuracy created by our in-house and low-cost LiDAR and camera integrated device. The closed-loop error provided by our SLAM system with and without this enhancement is 1.03m and 18.28m respectively. The scale estimation of the monocular visual SLAM is also significantly improved (0.01 versus 0.98). In addition, a novel camera-LiDAR calibration workflow is also provided for large-scale 3D mapping. This paper demonstrates the application and research potential of deep-learning based vision SLAM with image and LiDAR sensors.

Cross-view SLAM solver: Global pose estimation of monocular ground-level video frames for 3D reconstruction using a reference 3D model from satellite images

Accurate pose estimation of monocular ground-level images with respect to a satellite/aerial photogrammetric dataset is an extremely challenging task. Existing solutions often perform an offline post-registration on 3D results from both sources, which however, suffers from non-rigid geometric distortions of the 3D monocular reconstruction and the lack of overlaps between the air to ground content. This paper provides an online solution that performs accurate pose estimation of the ground images with respect to a 3D model derived from satellite images, followed by a dense 3D reconstruction. Our solution takes advantage of the simultaneous localization and mapping (SLAM) paradigm to dynamically incorporate reference observations from the satellite 3D model during the incremental pose estimation, called a cross-view SLAM solver, which leverages both ground-to-satellite error and image-level reprojection errors at the frame level to yield image poses that are well-registered to the satellite 3D model for facade point cloud reconstruction. This process also has the advantage of correcting non-rigid distortions and trajectory drifts that are often presented in monocular SLAM systems. In addition, our solution leverages both the geometric and semantic information from the satellite model and ground images to perform a per-frame correction for frame-level pose initialization, in which a novel scheme called pose buffer is introduced to initialize the pose of each keyframe through robust visual hull alignment of ground objects. The proposed approach has been experimented using four trajectories of monocular videos collections (around 7,000 frames per trajectory on average) and a 3D semantic model from multi-view satellite images to estimate the poses of the video frames and yield point clouds consistent with the satellite 3D models, evaluated by using LiDAR ground-truth. Both qualitative and quantitative experiments demonstrate that our solution yields accurate, drift-free poses and point clouds consistent with the satellite data and visually much more pleasing 3D models with facade information. Compared to the LiDAR ground-truth, the derived 3D models with ground-level images have achieved a mean absolute error of 1.78 m (improved from 3.15 m achieved using SLAM without utilizing satellite 3D models) (A testing program will be made available through https://github.com/GDAOSU/Cross-View-SLAM).

SO-SLAM: Semantic Object SLAM With Scale Proportional and Symmetrical Texture Constraints

Object SLAM introduces the concept of objects into Simultaneous Localization and Mapping (SLAM) and helps understand indoor scenes for mobile robots and object-level interactive applications. The state-of-art object SLAM systems face challenges such as partial observations, occlusions, unobservable problems, limiting the mapping accuracy and robustness. This letter proposes a novel monocular Semantic Object SLAM (SO-SLAM) system that addresses the introduction of object spatial constraints. We explore three representative spatial constraints, including scale proportional constraint, symmetrical texture constraint and plane supporting constraint. Based on these semantic constraints, we propose two new methods - a more robust object initialization method and an orientation fine optimization method. We have verified the performance of the algorithm on the public datasets and an author-recorded mobile robot dataset and achieved a significant improvement on mapping effects. We will release the code here. <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup>

A Mono SLAM Method Based on Depth Estimation by DenseNet-CNN

Currently, SLAM (simultaneous localization and mapping) systems based on monocular cameras cannot directly obtain depth information, and most of them have problems with scale uncertainty and need to be initialized. In some application scenarios that require navigation and obstacle avoidance, the inability to achieve dense mapping is also a defect of monocular SLAM. In response to the above problems, this paper proposes a method which learns depth estimation by DenseNet and CNN for a monocular SLAM system. We use an encoder-decoder architecture based on transfer learning and convolutional neural networks to estimate the depth information of monocular RGB images. At the same time, through the front-end ORB feature extraction and the back-end direct RGB-D Bundle Adjustment optimization method, it is possible to obtain accurate camera poses and achieve dense indoor mapping when using estimated depth information. The experimental results show that the monocular depth estimation model used in this paper can achieve good results, and it is also competitive in comparison with the current popular methods. On this basis, the error of camera pose estimation is also smaller than traditional monocular SLAM solutions and can complete the dense indoor reconstruction task. It is a complete SLAM system based on monocular camera.

3FO: The Three-Frame-Only Approach for Fast and Accurate Monocular SLAM Initialization

The monocular simultaneous localization and mapping (SLAM) system is one of the most important among all visual or visual-inertial SLAM (VSLAM or VISLAM) systems due to its low cost, easy calibration and identification. Initialization is always crucial to bootstrap the monocular SLAM system. With the rapid growth of some quick-start required SLAM applications, e.g., augmented reality (AR) and unmanned aerial vehicles (UAVs), devising faster and more accurate initialization has become a central problem. Traditional initialization uses the first two frames to create landmarks and then computes camera poses for the subsequential frames, using a 3D-to-2D perspective-n-points (PnP) mechanism. In this paper, we propose a novel three-frame-only (3FO) initialization approach for the monocular SLAM system, which consists of two steps. In the first step, we use the first two frames to preinitialize poses and landmarks, and in the second step, we use the second and third frames to improve the preinitialization by using the scale consistency of the landmarks generated in the first step to filter out outliers and using inliers to generate more robust landmarks. In both steps, we use the pretrained multilayer perceptron (MLP) combined with homotopy continuation to solve the essential matrices. Finally, we perform a global bundle adjustment (BA) to refine the three camera poses and all the created landmarks. The proposed 3FO initialization approach is evaluated experimentally on the EuRoC benchmark data set with the initialization time and trajectory metrics. The results show that, compared to the traditional ORB-SLAM2 initialization, the 3FO approach reduces the initialization time by 5 times and improves the accuracy by 36.7% on average.

딥 러닝을 이용한 단일 카메라 SLAM에서 스케일 드리프트 감소

Monocular SLAM(Simultaneous Localization and Mapping) systems have such advantages as low cost and light weight compared to stereo or laser range finder based SLAM systems. However they also have relatively large errors when they measure the distances between a vehicle and some distinct objects, which can lead to scale ambiguity or scale drift. In this paper, we suggest a scale drift-aware monocular SLAM system using CNN(Convolutional Neural Network) which is one of the key technologies of Deep Learning. CNN nowadays has proved its performances especially in computer vision. We have trained the system with plenty of images of some predetermined objects using CNN. And then we can measure the absolute distances between a vehicle and already known objects and resolve the scale drift problems. The suggested system has been evaluated by several experiments.

Improved Point-Line Feature Based Visual SLAM Method for Complex Environments.

Traditional visual simultaneous localization and mapping (SLAM) systems rely on point features to estimate camera trajectories. However, feature-based systems are usually not robust in complex environments such as weak textures or obvious brightness changes. To solve this problem, we used more environmental structure information by introducing line segments features and designed a monocular visual SLAM system. This system combines points and line segments to effectively make up for the shortcomings of traditional positioning based only on point features. First, ORB algorithm based on local adaptive threshold was proposed. Subsequently, we not only optimized the extracted line features, but also added a screening step before the traditional descriptor matching to combine the point features matching results with the line features matching. Finally, the weighting idea was introduced. When constructing the optimized cost function, we allocated weights reasonably according to the richness and dispersion of features. Our evaluation on publicly available datasets demonstrated that the improved point-line feature method is competitive with the state-of-the-art methods. In addition, the trajectory graph significantly reduced drift and loss, which proves that our system increases the robustness of SLAM.

Spatial semantic graph enhanced monocular SLAM System

AbstractMore and more applications require accurate estimation of objects' poses. For example, Augmented Reality (AR) needs that for interactions among virtual and real objects. However, recent object‐level Simultaneous Localization and Mapping (SLAM) systems care more about trajectories than poses of objects. So, in this paper, we present SSG SLAM: a monocular SLAM system based on Spatial Semantic Graph (SSG), leveraging spatial relations to achieve good estimation of objects' poses. First, we put forward the design of SSG to organize objects and their spatial relations. Then, SSG is utilized in a monocular object SLAM and new constraints are proposed based on spatial relations. Finally, experimental evaluations performed on SSG‐Dataset show that our approach outperforms the baseline object SLAM system, as well as providing visual consistency in an AR demo.

Multirobot Collaborative Monocular SLAM Utilizing Rendezvous

Multirobot simultaneous localization and mapping (SLAM) requires technical ingredients such as systematic construction of multiple SLAM systems and collection of information from each robot. In particular, map fusion is an essential process of multirobot SLAM that combines multiple local maps estimated by team robots into a global map. Fusion of multiple local maps is usually based on interloop detection that recognizes the same scene visited by multiple robots, or robot rendezvous where a member(s) of a robot team is observed in another member's images. This article proposes a collaborative monocular SLAM including a map fusion algorithm that utilizes rendezvous, which can happen when multirobot team members operate in close proximity. Unlike existing rendezvous-based approaches that require additional sensors, the proposed system uses a monocular camera only. Our system can recognize robot rendezvous using nonstatic features (NSFs) without fiducial markers to identify team robots. NSFs, which are abandoned as outliers in typical SLAM systems for not supporting ego-motion, can include relative bearing measurements between robots in a rendezvous situation. The proposed pipeline consists of the following: first, a feature identification module that extracts the relative bearing measurements between robots from NSFs consisting of anonymous bearing vectors with false positives, and second, a map fusion module that integrates the map from the observer robot with the maps from the observed robots using identified relative measurements. The feature identification module can operate quickly using the proposed alternating minimization algorithm formulated by two subproblems with closed-form solutions. The experimental results confirm that our collaborative monocular SLAM system recognizes rendezvous rapidly and robustly, and fuses local maps of team robots into a global map accurately.

Robust Visual-Inertial Navigation System for Low Precision Sensors under Indoor and Outdoor Environments

Simultaneous Localization and Mapping (SLAM) has always been the focus of the robot navigation for many decades and becomes a research hotspot in recent years. Because a SLAM system based on vision sensor is vulnerable to environment illumination and texture, the problem of initial scale ambiguity still exists in a monocular SLAM system. The fusion of a monocular camera and an inertial measurement unit (IMU) can effectively solve the scale blur problem, improve the robustness of the system, and achieve higher positioning accuracy. Based on a monocular visual-inertial navigation system (VINS-mono), a state-of-the-art fusion performance of monocular vision and IMU, this paper designs a new initialization scheme that can calculate the acceleration bias as a variable during the initialization process so that it can be applied to low-cost IMU sensors. Besides, in order to obtain better initialization accuracy, visual matching positioning method based on feature point is used to assist the initialization process. After the initialization process, it switches to optical flow tracking visual positioning mode to reduce the calculation complexity. By using the proposed method, the advantages of feature point method and optical flow method can be fused. This paper, the first one to use both the feature point method and optical flow method, has better performance in the comprehensive performance of positioning accuracy and robustness under the low-cost sensors. Through experiments conducted with the EuRoc dataset and campus environment, the results show that the initial values obtained through the initialization process can be efficiently used for launching nonlinear visual-inertial state estimator and positioning accuracy of the improved VINS-mono has been improved by about 10% than VINS-mono.

DeepSLAM: A Robust Monocular SLAM System With Unsupervised Deep Learning

In this article, we propose DeepSLAM, a novel unsupervised deep learning based visual simultaneous localization and mapping (SLAM) system. The DeepSLAM training is fully unsupervised since it only requires stereo imagery instead of annotating ground-truth poses. Its testing takes a monocular image sequence as the input. Therefore, it is a monocular SLAM paradigm. DeepSLAM consists of several essential components, including Mapping-Net, Tracking-Net, Loop-Net, and a graph optimization unit. Specifically, the Mapping-Net is an encoder and decoder architecture for describing the 3-D structure of environment, whereas the Tracking-Net is a recurrent convolutional neural network architecture for capturing the camera motion. The Loop-Net is a pretrained binary classifier for detecting loop closures. DeepSLAM can simultaneously generate pose estimate, depth map, and outlier rejection mask. In this article, we evaluate its performance on various datasets, and find that DeepSLAM achieves good performance in terms of pose estimation accuracy, and is robust in some challenging scenes.