ICRA 2023 Papers

This paper proposes a set of tactile based skills to perform robotic cable routing operations for deformable linear objects (DLOs) characterized by considerable stiffness and constrained at both ends. In particular, tactile data are exploited to reconstruct the shape of the grasped portion of the DLO and to estimate the future local one. This information is exploited to obtain a grasping configuration aligned to the local shape of the DLO, starting from a rough initial grasping pose, and to follow the DLO's contour in the three-dimensional space. Taking into account the distance travelled along the arc length of the DLO, the robot can detect the cable segments that must be firmly grasped and inserted in intermediate clips, continuing then to slide along the contour until the next DLO's portion, that has to be clipped, is reached. The proposed skills are experimentally validated with an industrial robot on different DLOs in several configurations and on a cable routing use case.

Humans can effortlessly perform very complex, dexterous manipulation tasks by reacting to sensor observations. In contrast, robots can not perform reactive manipulation and they mostly operate in open-loop while interacting with their environment. Consequently, the current manipulation algorithms either are inefficient in performance or can only work in highly structured environments. In this paper, we present closed-loop control of a complex manipulation task where a robot uses a tool to interact with objects. Manipulation using a tool leads to complex kinematics and contact constraints that need to be satisfied for generating feasible manipulation trajectories. We first present an open-loop controller design using Non-Linear Programming (NLP) that satisfies these constraints. In order to design a closed-loop controller, we present a pose estimator of objects and tools using tactile sensors. Using our tactile estimator, we design a closed-loop controller based on Model Predictive Control (MPC). The proposed algorithm is verified using a 6 DoF manipulator on tasks using a variety of objects and tools. We verify that our closed-loop controller can successfully perform tool manipulation under several unexpected contacts.

With the goal of providing an alternative to optical and other tactile sensors, we set out to stress test the object shape identification capabilities of barometric tactile arrays in robotic manipulation tasks. These sensors are superior to optical devices in terms of form factor, ease of fabrication, and data reading/processing speeds, but lack the necessary spatial resolution to identify surface shapes via a single contact. To compensate, we utilize in-hand-manipulation, specifically in- hand-rolling to identify object shapes via a spatiotemporal approach. To increase task difficulty, we only use three neighboring barometric sensors and designed strict experiment requirements with the purpose of creating a set of extremely confusable test objects. The E- TRoll robotic hand, equipped with a barometric tactile array on one finger, was used to roll test objects within its grasp, taking just under 3.4 seconds for data collection under the fastest tested speed setting, compared to 33 seconds in our previous work. We also designed and implemented a feature extraction algorithm, based and improved upon our recently published algorithm. This captures enough information from the collected spatiotemporal data samples for successful classification with only 13 features. Finally, a bagged tree classification algorithm was trained and optimized with data from 1,164 trials of rolling 9 prismatic test objects, leading to a five-fold cross validation accuracy of 90.5% for identifying the 9 object classes.

Radar sensors employed for environment perception, e.g. in autonomous vehicles, output a lot of unwanted clutter. These points, for which no corresponding real objects exist, are a major source of errors in following processing steps like object detection or tracking. We therefore present two novel neural network setups for identifying clutter. The input data, network architectures and training configuration are adjusted specifically for this task. Special attention is paid to the downsampling of point clouds composed of multiple sensor scans. In an extensive evaluation, the new setups display substantially better performance than existing approaches. Because there is no suitable public data set in which clutter is annotated, we design a method to automatically generate the respective labels. By applying it to existing data with object annotations and releasing its code, we effectively create the first freely available radar clutter data set representing realworld driving scenarios. Code and instructions are accessible at www.github.com/kopp-j/clutter-ds.

Tactile pose estimation and tactile servoing are fundamental capabilities of robot touch. Reliable and precise pose estimation can be provided by applying deep learning models to high-resolution optical tactile sensors. Given the recent successes of Graph Neural Network (GNN) and the effectiveness of Voronoi features, we developed a Tactile Voronoi Graph Neural Network (Tac-VGNN) to achieve reliable pose-based tactile servoing relying on a biomimetic optical tactile sensor (TacTip). The GNN is well suited to modeling the distribution relationship between shear motions of the tactile markers, while the Voronoi diagram supplements this with area-based tactile features related to contact depth. The experiment results showed that the Tac-VGNN model can help enhance data interpretability during graph generation and model training efficiency significantly than CNN-based methods. It also improved pose estimation accuracy along vertical depth by 28.57% over vanilla GNN without Voronoi features and achieved better performance on the real surface following tasks with smoother robot control trajectories. For more project details, please view our website: https://sites.google.com/view/tac-vgnn/home

Distance estimation from vision is fundamental for a myriad of robotic applications such as navigation, manipu-lation, and planning. Inspired by the mammal's visual system, which gazes at specific objects, we develop two novel constraints relating time-to-contact, acceleration, and distance that we call the \tau\tau -constraint and \Phi\Phi -constraint. They allow an active (moving) camera to estimate depth efficiently and accurately while using only a small portion of the image. The constraints are applicable to range sensing, sensor fusion, and visual servoing. We successfully validate the proposed constraints with two experiments. The first applies both constraints in a trajectory estimation task with a monocular camera and an Inertial Measurement Unit (IMU). Our methods achieve 30-70% less average trajectory error while running 25\times25\times and 6.2\times6.2\times faster than the popular Visual-Inertial Odometry methods VINS-Mono and ROVIO respectively. The second experiment demonstrates that when the constraints are used for feedback with efference copies the resulting closed loop system's eigenvalues are invariant to scaling of the applied control signal. We believe these results indicate the \tau\tau and \Phi\Phi constraint's potential as the basis of robust and efficient algorithms for a multitude of robotic applications.

We propose TRADE for robust tracking and 3D localization of a moving target in complex environments, from UAVs equipped with a single camera. Ultimately TRADE enables 3d-aware target following. Tracking-by-detection approaches are vulnerable to target switching, especially between similar objects. Thus, TRADE predicts and incorporates the target 3D trajectory to select the right target from the tracker's response map. Unlike static environments, depth estimation of a moving target from a single camera is an ill-posed problem. Therefore we propose a novel 3D localization method for ground targets on complex terrain. It reasons about scene geometry by combining ground plane segmentation, depth-from-motion and single-image depth estimation. The benefits of using TRADE are demonstrated as tracking robustness and depth accuracy on several dynamic scenes simulated in this work. Additionally, we demonstrate autonomous target following using a thermal camera by running TRADE on a quadcopter's board computer.

Coexisting with others and interacting in society implies sharing knowledge and attention about world objects, events, features, episodes, and even imagination or abstract ideas in time and space. Inspired by human phenomenological, cognitive and behavioral research, this work focuses on the study of joint attention (JA) for human-robot interaction (HRI), based on two main assumptions: a) the perception and representation of attention jointness constitute an isomorphic relation, and b) inspiration on dynamic neural fields (DNF) theory is a promising way to investigate contextual and non-linear spatio-temporal relations underlying attention and knowledge sharing in HRI. Taking into account the previous considerations, we propose a topology-based model for JA named TOP-JAM, which is able to represent and track in real-time JA states, from observations of behavioral data. More importantly, the model consists in a representation that can be directly understood by human beings, which conforms to robo-ethical principles in social robotics. This study evaluates computational properties of the model in simulation. Through a real experiment with the robot Pepper, the study shows that TOP-JAM is able to track JA in a triad interaction scenario.

In autonomous driving, an accurate understanding of environment, e.g., the vehicle-to-vehicle and vehicle-to-lane interactions, plays a critical role in many driving tasks such as trajectory prediction and motion planning. Environment information comes from high-definition (HD) map and historical trajectories of vehicles. Due to the heterogeneity of the map data and trajectory data, many data-driven models for trajectory prediction and motion planning extract vehicle-to-vehicle and vehicle-to-lane interactions in a separate and sequential manner. However, such a manner may capture biased interpretation of interactions, causing lower prediction and planning accuracy. Moreover, separate extraction leads to a complicated model structure and hence the overall efficiency and scalability are sacrificed. To address the above issues, we propose an environment representation, Temporal Occupancy Flow Graph (TOFG). Specifically, the occupancy flow-based representation unifies the map information and vehicle trajectories into a homogeneous data format and enables a consistent prediction. The temporal dependencies among vehicles can help capture the change of occupancy flow timely to further promote model performance. To demonstrate that TOFG is capable of simplifying the model architecture, we incorporate TOFG with a simple graph attention (GAT) based neural network and propose TOFG-GAT, which can be used for both trajectory prediction and motion planning. Experiment results show that TOFG-GAT achieves better or competitive performance than all the SOTA baselines with less training time.

Transparent objects are widely used in industrial automation and daily life. However, robust visual recognition and perception of transparent objects have always been a major challenge. Currently, most commercial-grade depth cameras are still not good at sensing the surfaces of transparent objects due to the refraction and reflection of light. In this work, we present a transformer-based transparent object depth estimation approach from a single RGB-D input. We observe that the global characteristics of the transformer make it easier to extract contextual information to perform depth estimation of transparent areas. In addition, to better enhance the fine-grained features, a feature fusion module (FFM) is designed to assist coherent prediction. Our empirical evidence demonstrates that our model delivers significant improvements in recent popular datasets, e.g., 25% gain on RMSE and 21% gain on REL compared to previous state-of-the-art convolutional-based counterparts in ClearGrasp dataset. Extensive results show that our transformer-based model enables better aggregation of the object's RGB and inaccurate depth information to obtain a better depth representation. Our code and the pre-trained model are available at https://github.com/yuchendoudou/TODE.

Aerial-aquatic vehicles are capable to move in the two most dominant fluids, making them more promising for a wide range of applications. We propose a prototype with special designs for propulsion and thruster configuration to cope with the vast differences in the fluid properties of water and air. For propulsion, the operating range is switched for the different mediums by the dual-speed propulsion unit, providing sufficient thrust and also ensuring output efficiency. For thruster configuration, thrust vectoring is realized by the rotation of the propulsion unit around the mount arm, thus enhancing the underwater maneuverability. This paper presents a quadrotor prototype of this concept and the design details and realization in practice.

Tactile recognition of 3D objects remains a challenging task. Compared to 2D shapes, the complex geometry of 3D surfaces requires richer tactile signals, more dexterous actions, and more advanced encoding techniques. In this work, we propose TANDEM3D, a method that applies a co-training framework for exploration and decision making to 3D object recognition with tactile signals. Starting with our previous work, which introduced a co-training paradigm for 2D recognition problems, we introduce a number of advances that enable us to scale up to 3D. TANDEM3D is based on a novel encoder that builds 3D object representation from contact positions and normals using PointNet++. Furthermore, by enabling 6DOF movement, TANDEM3D explores and collects discriminative touch information with high efficiency. Our method is trained entirely in simulation and validated with real-world experiments. Compared to state-of-the-art baselines, TANDEM3D achieves higher accuracy and a lower number of actions in recognizing 3D objects and is also shown to be more robust to different types and amounts of sensor noise.

A robot guide dog has compelling advantages over animal guide dogs for its cost-effectiveness, the potential for mass production, and low maintenance burden. However, despite the long history of guide dog robot research, previous studies were conducted with little or no consideration of how the guide dog handler and the guide dog work as a team for navigation. To develop a robotic guiding system that genuinely benefits blind or visually impaired individuals, we performed qualitative research, including interviews with guide dog handlers, trainers, and first-hand blindfold walking experiences with various guide dogs. We build a collaborative indoor navigation scheme for a guide dog robot that includes preferred features such as speed and directional control. For collaborative navigation, we propose a semantic-aware local path planner that enables safe and efficient guiding work by utilizing semantic information about the environment and considering the handler's position and directional cues to determine the collision-free path. We evaluate our integrated robotic system by testing blindfolded walking in indoor settings and demonstrate guide dog-like navigation behavior by avoiding obstacles at typical gait speed (0.7m/s). The following demonstration video link includes an audio description: https://youtu.be/YxlcMeaL7GA

Underwater image enhancement (UIE) is vital for high-level vision-related underwater tasks. Although learning-based UIE methods have made remarkable achievements in recent years, it's still challenging for them to consistently deal with various underwater conditions, which could be caused by: 1) the use of the simplified atmospheric image formation model in UIE may result in severe errors; 2) the network trained solely with synthetic images might have difficulty in generalizing well to real underwater images. In this work, we, for the first time, propose a framework SyreaNet for UIE that integrates both synthetic and real data under the guidance of the revised underwater image formation model and novel domain adaptation (DA) strategies. First, an underwater image synthesis module based on the revised model is proposed. Then, a physically guided disentangled network is designed to predict the clear images by combining both synthetic and real underwater images. The intra- and inter-domain gaps are abridged by fully exchanging the domain knowledge. Extensive experiments demonstrate the superiority of our framework over other state-of-the-art (SOTA) learning-based UIE methods qualitatively and quantitatively. The code and dataset are publicly available at https://github.com/RockWenJJ/SyreaNet.git.

Human action recognition is a challenging problem, particularly when there is high variability in factors such as subject appearance, backgrounds and viewpoint. While deep neural networks (DNNs) have been shown to perform well on action recognition tasks, they typically require large amounts of high-quality labeled data to achieve robust performance across a variety of conditions. Synthetic data has shown promise as a way to avoid the substantial costs and potential ethical concerns associated with collecting and labeling enormous amounts of data in the real-world. However, synthetic data may differ from real data in important ways. This phenomenon, known as domain shift, can limit the utility of synthetic data in robotics applications. To mitigate the effects of domain shift, substantial effort is being dedicated to the development of domain adaptation (DA) techniques. Yet, much remains to be understood about how best to develop these techniques. In this paper, we introduce a new dataset called Robot Control Gestures (RoCoG-v2). The dataset is composed of both real and synthetic videos from seven gesture classes, and is intended to support the study of synthetic-to-real domain shift for video-based action recognition. Our work expands upon existing datasets by focusing the action classes on gestures for human-robot teaming, as well as by enabling investigation of domain shift in both ground and aerial views. We present baseline results using state-of-the-art action recognition and domain adaptation algorithms and offer initial insight on tackling the synthetic-to-real and ground-to-air domain shifts. Instructions on accessing the dataset can be found at https://github.com/reddyav1/RoCoG-v2.

We study automated test generation for testing discrete decision-making modules in autonomous systems. Linear temporal logic is used to encode the system specification - requirements of the system under test - and the test specification, which is unknown to the system and describes the desired test behavior. The reactive test synthesis problem is to find constraints on system actions such that in a test execution, both the system and test specifications are satisfied. To do this, we use the specifications and their corresponding Büchi automata to construct the specification product automaton. Then, a virtual product graph representing all possible test executions of the system is constructed from the transition system and the specification product automaton. The main result of this paper is framing the test synthesis problem as a multi-commodity network flow optimization. This optimization is used to derive reactive constraints on system actions, which constitute the test environment. The resulting test environment ensures that the system meets the test specification while also satisfying the system specification. We illustrate this framework in simulation using grid world examples and demonstrate it on hardware with the Unitree A1 quadruped, where we test dynamic locomotion behaviors reactively.

Motivated by the goal of endowing robots with a means for focusing attention in order to operate reliably in complex, uncertain, and time-varying environments, we consider how a robot can (i) determine which portions of its environment to pay attention to at any given point in time, (ii) infer changes in context (e.g., task or environment dynamics), and (iii) switch its attention accordingly. In this work, we tackle these questions by modeling context switches in a time-varying Markov decision process (MDP) framework. We utilize the theory of bisimulation-based state abstractions in order to synthesize mechanisms for paying attention to context-relevant information. We then present an algorithm based on Bayesian inference for detecting changes in the robot's context (task or environment dynamics) as it operates online, and use this to trigger switches between different abstraction-based attention mechanisms. Our approach is demonstrated on two examples: (i) an illustrative discrete-state tracking problem, and (ii) a continuous-state tracking problem implemented on a quadrupedal hardware platform. These examples demonstrate the ability of our approach to detect context switches online and robustly ignore task-irrelevant distractors by paying attention to context-relevant information.

Monocular depth estimation plays a critical role in various computer vision and robotics applications such as localization, mapping, and 3D object detection. Recently, learning-based algorithms achieve huge success in depth estimation by training models with a large amount of data in a supervised manner. However, it is challenging to acquire dense ground truth depth labels for supervised training, and the unsupervised depth estimation using monocular sequences emerges as a promising alternative. Unfortunately, most studies on unsupervised depth estimation explore loss functions or occlusion masks, and there is little change in model architecture in that ConvNet-based encoder-decoder structure becomes a de-facto standard for depth estimation. In this paper, we employ a convolution-free Swin Transformer as an image feature extractor so that the network can capture both local geometric features and global semantic features for depth estimation. Also, we propose a Densely Cascaded Multi-scale Network (DCMNet) that connects every feature map directly with another from different scales via a top-down cascade pathway. This densely cascaded connectivity reinforces the interconnection between decoding layers and produces high-quality multi-scale depth outputs. The experiments on two different datasets, KITTI and Make3D, demonstrate that our proposed method outperforms existing state-of-the-art unsupervised algorithms.

Accurate self and relative state estimation are the critical preconditions for completing swarm tasks, e.g., collaborative autonomous exploration, target tracking, search and rescue. This paper proposes Swarm-LIO: a fully decentralized state estimation method for aerial swarm systems, in which each drone performs precise ego-state estimation, exchanges ego-state and mutual observation information by wireless communication, and estimates relative state with respect to (w.r.t.) the rest of UAVs, all in real-time and only based on LiDAR-inertial measurements. A novel 3D LiDAR-based drone detection, identification and tracking method is proposed to obtain observations of teammate drones. The mutual observation measurements are then tightly-coupled with IMU and LiDAR measurements to perform real-time and accurate estimation of ego-state and relative state jointly. Extensive real-world experiments show the broad adaptability to complicated scenarios, including GPS-denied scenes, degenerate scenes for camera (dark night) or LiDAR (facing a single wall). Compared with ground-truth provided by motion capture system, the result shows the centimeter-level localization accuracy which outperforms other state-of-the-art LiDAR-inertial odometry for single UAV system.