ICRA 2021 论文列表

Reliable robot pose estimation is a key building block of many robot autonomy pipelines, with LiDAR localization being an active research domain. In this work, a versatile self-supervised LiDAR odometry estimation method is presented, in order to enable the efficient utilization of all available LiDAR data while maintaining real-time performance. The proposed approach selectively applies geometric losses during training, being cognizant of the amount of information that can be extracted from scan points. In addition, no labeled or ground-truth data is required, hence making the presented approach suitable for pose estimation in applications where accurate ground-truth is difficult to obtain. Furthermore, the presented network architecture is applicable to a wide range of environments and sensor modalities without requiring any network or loss function adjustments. The proposed approach is thoroughly tested for both indoor and outdoor real-world applications through a variety of experiments using legged, tracked and wheeled robots, demonstrating the suitability of learning-based LiDAR odometry for complex robotic applications.

Deep learning is the essential building block of state-of-the-art person detectors in 2D range data. However, only a few annotated datasets are available for training and testing these deep networks, potentially limiting their performance when deployed in new environments or with different LiDAR models. We propose a method, which uses bounding boxes from an image-based detector (e.g. Faster R-CNN) on a calibrated camera to automatically generate training labels (called pseudo-labels) for 2D LiDAR-based person detectors. Through experiments on the JackRabbot dataset with two detector models, DROW3 and DR-SPAAM, we show that self-supervised detectors, trained or fine-tuned with pseudolabels, outperform detectors trained only on a different dataset. Combined with robust training techniques, the self-supervised detectors reach a performance close to the ones trained using manual annotations of the target dataset. Our method is an effective way to improve person detectors during deployment without any additional labeling effort, and we release our source code to support relevant robotic applications.

In this paper, we present self-supervised shared latent embedding (S3LE), a data-driven motion retargeting method that enables the generation of natural motions in humanoid robots from motion capture data or RGB videos. While it requires paired data consisting of human poses and their corresponding robot configurations, it significantly alleviates the necessity of time-consuming data-collection via novel paired data generating processes. Our self-supervised learning procedure consists of two steps: automatically generating paired data to bootstrap the motion retargeting, and learning a projection-invariant mapping to handle the different expressivity of humans and humanoid robots. Furthermore, our method guarantees that the generated robot pose is collision-free and satisfies position limits by utilizing nonparametric regression in the shared latent space. We demonstrate that our method can generate expressive robotic motions from both the CMU motion capture database and YouTube videos.

We present a self-supervised learning approach for the semantic segmentation of lidar frames. Our method is used to train a deep point cloud segmentation architecture without any human annotation. The annotation process is automated with the combination of simultaneous localization and mapping (SLAM) and ray-tracing algorithms. By performing multiple navigation sessions in the same environment, we are able to identify permanent structures, such as walls, and disentangle short-term and long-term movable objects, such as people and tables, respectively. New sessions can then be performed using a network trained to predict these semantic labels. We demonstrate the ability of our approach to improve itself over time, from one session to the next. With semantically filtered point clouds, our robot can navigate through more complex scenarios, which, when added to the training pool, help to improve our network predictions. We provide insights into our network predictions and show that our approach can also improve the performances of common localization techniques.

Self-supervised learning algorithms that compute depth map from monocular videos have achieved remarkable performance on urban scenes and have been applied extensively. These techniques still face significant challenges, however, when applied directly to endoscopic videos because of the brightness variations from frame to frame and inadequate representation learning during the training phase. Inspired by the optical flow for motion alignment between adjacent frames, we design a AFNet with structural stability loss and residual-based smoothness loss to learn the appearance flow across adjacent frames, which handles the brightness inconsistency issue efficaciously. In addition, we propose a novel self-attention mechanism named feature scaling module to alleviate the inadequate representation learning problem. In a comparison study to the current state-of-the-art self-supervised methods explored for urban videos on the SCARED dataset, the developed model surpasses existing methods by a large margin.

Several animal species self-organize into large groups to leverage vital behaviors such as foraging, construction, or predator evasion. With the advancement of robotics and automation, engineered multi-agent systems have been inspired to achieve similarly high degrees of scalable, robust, and adaptable autonomy through decentralized and dynamic coordination. So far however, they have been most successfully demonstrated above ground or with partial assistance from central controllers and external tracking. Here we demonstrate an underwater robot collective that realizes full spatiotemporal coordination. Using the example of fish-inspired evasive maneuvers, our robots display alignment, formation control, and coordinated escape, enabled by real-time on-board multi-robot tracking and local decision making. Accompanied by a custom simulator, this robotic platform advances the physically- validated development of algorithms for collective behaviors and future applications including collective exploration, tracking and capture, or environmental sampling.

Imitation learning (IL) enables robots to acquire skills quickly by transferring expert knowledge, which is widely adopted in reinforcement learning (RL) to initialize exploration. However, in long-horizon motion planning tasks, a challenging problem in deploying IL and RL methods is how to generate and collect massive, broadly distributed data such that these methods can generalize effectively. In this work, we solve this problem using our proposed approach called self-imitation learning by planning (SILP), where demonstration data are collected automatically by planning on the visited states from the current policy. SILP is inspired by the observation that successfully visited states in the early reinforcement learning stage are collision-free nodes in the graph-search based motion planner, so we can plan and relabel robot's own trials as demonstrations for policy learning. Due to these self-generated demonstrations, we relieve the human operator from the laborious data preparation process required by IL and RL methods in solving complex motion planning tasks. The evaluation results show that our SILP method achieves higher success rates and enhances sample efficiency compared to selected baselines, and the policy learned in simulation performs well in a real-world placement task with changing goals and obstacles.

Depth completion aims at inferring a dense depth image from sparse depth measurement since glossy, transparent or distant surface cannot be scanned properly by the sensor. Most of existing methods directly interpolate the missing depth measurements based on pixel-wise image content and the corresponding neighboring depth values. Consequently, this leads to blurred boundaries or inaccurate structure of object. To address these problems, we propose a novel self-guided instance-aware network (SG-IANet) that: (1) utilize self-guided mechanism to extract instance-level features that is needed for depth restoration, (2) exploit the geometric and context information into network learning to conform to the underlying constraints for edge clarity and structure consistency, (3) regularize the depth estimation and mitigate the impact of noise by instance-aware learning, and (4) train with synthetic data only by domain randomization to bridge the reality gap. Extensive experiments on synthetic and real world dataset demonstrate that our proposed method outperforms previous works. Further ablation studies give more insights into the proposed method and demonstrate the generalization capability of our model.

We consider the problem of security-aware planning in an unknown stochastic environment, in the presence of attacks on control signals (i.e., actuators) of the robot. We model the attacker as an agent who has the full knowledge of the controller as well as the employed intrusion-detection system and who wants to prevent the controller from performing tasks while staying stealthy. We formulate the problem as a stochastic game between the attacker and the controller and present an approach to express the objective of such an agent and the controller as a combined linear temporal logic (LTL) formula. We then show that the planning problem, described formally as the problem of satisfying an LTL formula in a stochastic game, can be solved via model-free reinforcement learning when the environment is completely unknown. Finally, we illustrate and evaluate our methods on two robotic planning case studies.

Search-based methods that use motion primitives can incorporate the system’s dynamics into the planning and thus generate dynamically feasible MAV trajectories that are globally optimal. However, searching high-dimensional state lattices is computationally expensive. Local multiresolution is a commonly used method to accelerate spatial path planning. While paths within the vicinity of the robot are represented at high resolution, the representation gets coarser for more distant parts. In this work, we apply the concept of local multiresolution to high-dimensional state lattices that include velocities and accelerations. Experiments show that our proposed approach significantly reduces planning times. Thus, it increases the applicability to large dynamic environments, where frequent replanning is necessary.

Path planning for robotic coverage is the task of determining a collision-free robot trajectory that observes all points of interest in an environment. Robots employed for such tasks are often capable of exercising active control over onboard observational sensors during navigation. We address the problem of planning robot and sensor trajectories that maximize information gain in such tasks, where the robot needs to cover points of interest with its sensor footprint. Search-based planners in general guarantee completeness and provable bounds on sub-optimality with respect to an underlying graph discretization. However, searching for kinodynamically feasible paths in the joint space of robot and sensor state variables with standard search is computationally expensive. We propose two alternative search-based approaches to this problem. The first solves for robot and sensor trajectories independently in decoupled state spaces while maintaining a history of sensor headings during the search. The second is a two-step approach that first quickly computes a solution in decoupled state spaces and then refines it by searching its local neighborhood in the joint space for a better solution. We evaluate our approaches in simulation with a kinodynamically constrained unmanned aerial vehicle performing coverage over a 2D environment and show their benefits.

This paper presents a search-based partial motion planner for generating feasible trajectories of car-like robots in highly dynamic environments. The planner searches for smooth, safe, and near-time-optimal trajectories by exploring a state graph built on motion primitives. To enable fast online planning, we propose an efficient path searching algorithm based on the aggregation and pruning of motion primitives. We then propose a fast collision checking algorithm that takes into account the motions of moving obstacles. The algorithm linearizes relative motions between the robot and obstacles, and then checks collisions by calculating a point-line distance. Benefiting from the fast searching and collision checking algorithms, the planner can effectively explore the state-time space to generate near-time-optimal solutions. Experiments show that the proposed method can generate feasible trajectories within milliseconds while maintaining a higher success rate than up-to-date methods, which significantly demonstrates its advantages.

Robots in human environments will need to interact with a wide variety of articulated objects such as cabinets, drawers, and dishwashers while assisting humans in performing day-to-day tasks. Existing methods either require objects to be textured or need to know the articulation model category a priori for estimating the model parameters for an articulated object. We propose ScrewNet, a novel approach that estimates an object’s articulation model directly from depth images without requiring a priori knowledge of the articulation model category. ScrewNet uses screw theory to unify the representation of different articulation types and perform category-independent articulation model estimation. We evaluate our approach on two benchmarking datasets and three real-world objects and compare its performance with a current state-of-the-art method. Results demonstrate that ScrewNet can successfully estimate the articulation models and their parameters for novel objects across articulation model categories with better on average accuracy than the prior state-of-the-art method.

Soft robotic manipulators have been created and investigated for a number of applications due to their advantages over rigid robots. In minimally invasive surgery, for instance, soft robots have successfully demonstrated a number of benefits due to the compliant and flexible nature of the material they are made of. However, these type of robots struggle with performing tasks that require on-demand stiffness i.e. exerting higher forces to the surrounding environment. A number of semi-active and active mechanisms have been investigated to change and control the stiffness of soft robotic manipulators. Embedding these mechanisms in soft manipulators for spacerestricted applications can be challenging though.To better understand the inherent passive stiffness properties of soft manipulators, we propose a screw theory-based stiffness analysis for fluidic-driven continuum soft robotic manipulators. First, we derive the forward kinematics based on a parameter-based piece-wise constant curvature model. It is worth noting, our stiffness analysis can be conducted based on any freespace forward kinematic model. Then our stiffness analysis and mapping methodology is conducted based on screw theory. Initial results of our approach demonstrate the feasibility comparing computational and experimental data.

Grasping objects is one of the most important abilities that a robot needs to master in order to interact with its environment. Current state-of-the-art methods rely on deep neural networks trained to jointly predict a graspability score together with a regression of an offset with respect to grasp reference parameters. However, these two predictions are performed independently, which can lead to a decrease in the actual graspability score when applying the predicted offset. Therefore, in this paper, we extend a state-of-the-art neural network with a scorer that evaluates the graspability of a given position, and introduce a novel loss function which correlates regression of grasp parameters with graspability score. We show that this novel architecture improves performance from 82.13% for a state-of-the-art grasp detection network to 85.74% on Jacquard dataset. When the learned model is transferred onto a real robot, the proposed method correlating graspability and grasp regression achieves a 92.4% rate compared to 88.1% for the baseline trained without the correlation.

This paper presents a novel approach to recover outdated cross-covariance between correlated agents at the moment they perform joint observations. This allows to render Collaborative State Estimation (CSE) fully distributed, with communication only required for the moment of joint observation and most importantly, it significantly reduces the maintenance effort in case of high frequent propagation sensors. These properties make the approach suitable to a wide range of multi-robot applications. In our evaluation on a Quaternion-based Error-State Extended Kalman Filter (Q-ESEKF) using an Inertial Measurement Unit (IMU) as propagation sensor at a rate of 200Hz, we showed a significant speedup against our previous approach for maintaining a couple of interdependence. We compared the approach in total against four different approaches on both, a simulation and on a real-world dataset for Micro Aerial Vehicles (MAVs). Video: https://youtu.be/xkljfwbhMP0

This paper presents a novel policy representation for partially observable Markov decision processes (POMDPs) called circulant controllers and a provably efficient gradient-based algorithm for them. A formal mathematical description is provided that leverages circulant matrices for the controller’s stochastic node transitions. This structure is particularly effective for capturing decision-making patterns found in real-world domains with repeated periodic behaviors that adapt their cycles based on observation. This includes domains such as bipedal walking over varied terrain, pick-and-place tasks in warehouses, and home healthcare monitoring and medicine delivery in household environments. A performant gradient-based algorithm is presented with a detailed theoretical analysis, formally proving the algorithm’s improved performance, as well as circulant controllers’ structural properties. Experiments on these domains demonstrate that the proposed controller algorithm outperforms other state-of-the-art POMDP controller algorithms. The proposed novel controller approach is demonstrated on an actual robot performing a navigation task in a real household environment.

Autonomous systems like aircraft and assistive robots often operate in scenarios where guaranteeing safety is critical. Methods like Hamilton-Jacobi reachability can provide guaranteed safe sets and controllers for such systems. However, often these same scenarios have unknown or uncertain environments, system dynamics, or predictions of other agents. As the system is operating, it may learn new knowledge about these uncertainties and should therefore update its safety analysis accordingly. However, work to learn and update safety analysis is limited to small systems of about two dimensions due to the computational complexity of the analysis. In this paper we synthesize several techniques to speed up computation: decomposition, warm-starting, and adaptive grids. Using this new framework we can update safe sets by one or more orders of magnitude faster than prior work, making this technique practical for many realistic systems. We demonstrate our results on simulated 2D and 10D near-hover quadcopters operating in a windy environment.

This paper presents a novel multi-robot coverage path planning (CPP) algorithm - aka SCoPP - that provides a time-efficient solution, with workload balanced plans for each robot in a multi-robot system, based on their initial states. This algorithm accounts for discontinuities (e.g., no-fly zones) in a specified area of interest, and provides an optimized ordered list of way-points per robot using a discrete, computationally efficient, nearest neighbor path planning algorithm. This algorithm involves five main stages, which include the transformation of the user’s input as a set of vertices in geographical coordinates, discretization, load-balanced partitioning, auctioning of conflict cells in a discretized space, and a path planning procedure. To evaluate the effectiveness of the primary algorithm, a multi-unmanned aerial vehicle (UAV) post-flood assessment application is considered, and the performance of the algorithm is tested on three test maps of varying sizes. Additionally, our method is compared with a state-of-the-art method created by Guasella et al. Further analyses on scalability and computational time of SCoPP are conducted. The results show that SCoPP is superior in terms of mission completion time; its computing time is found to be under 2 mins for a large map covered by a 150-robot team, thereby demonstrating its computationally scalability.