Masayuki Inaba

Kazuki Sugihara, Moju Zhao, Takuzumi Nishio et al.

In recent years, multimodal locomotion capabilities have enabled robots to maneuver in both terrestrial and aerial domains. However, most of these robots are designed only for locomotion, and few possess the manipulation capabilities required for practical tasks. By adding a manipulator, ground robots can perform manipulation, and some drones with robotic arms have demonstrated aerial manipulation. Nonetheless, such multirotors cannot be directly used for manipulation on the ground, and this configuration itself is unsuitable for air-ground hybrid locomotion. This is because their thruster-centralized structure makes it difficult to achieve both sufficient degrees of freedom (DoF) for manipulation and stable motion with contact and transformation. Therefore, in this work, we develop a new multilink multirotor with thrusters on each link and capable of contact with the environments. This robot can perform terrestrial rolling locomotion, aerial flight locomotion, and manipulation in multiple environments using joint actuation. First, we introduce a minimal configuration design of the proposed robot. We also describe a kinematic model and propose a design for each component based on this model. Second, we propose a real-time control method based on nonlinear optimization that considers contact and joint motion, which can be applied to various multirotors. Third, we propose motion strategies that include contact constraints specific to air-ground hybrid multilink multirotors, and analyze the limitations of manipulation capabilities based on multi-contact model. Finally, we demonstrate a variety of motions in both domains using the implemented prototype. To the best of our knowledge, this is the first demonstration of air-ground hybrid locomotion and manipulation by a multilink multirotor.

Shumpei Wakabayashi, Kento Kawaharazuka, Kei Okada et al.

Robots are expected to manipulate objects in a safe and dexterous way. For example, washing dishes is a dexterous operation that involves scrubbing the dishes with a sponge and rinsing them with water. It is necessary to learn it safely without splashing water and without dropping the dishes. In this study, we propose a safe and dexterous manipulation system. The robot learns a dynamics model of the object by estimating the state of the object and the robot itself, the control input, and the amount of human assistance required (assistance rate) after the human corrects the initial trajectory of the robot's hands by interruptive direct teaching. By backpropagating the error between the estimated and the reference value using the acquired dynamics model, the robot can generate a control input that approaches the reference value, for example, so that human assistance is not required and the dish does not move excessively. This allows for adaptive rinsing and scrubbing of dishes with unknown shapes and properties. As a result, it is possible to generate safe actions that require less human assistance.

Kento Kawaharazuka, Yoshiki Obinata, Naoaki Kanazawa et al.

Various robot navigation methods have been developed, but they are mainly based on Simultaneous Localization and Mapping (SLAM), reinforcement learning, etc., which require prior map construction or learning. In this study, we consider the simplest method that does not require any map construction or learning, and execute open-vocabulary navigation of robots without any prior knowledge to do this. We applied an omnidirectional camera and pre-trained vision-language models to the robot. The omnidirectional camera provides a uniform view of the surroundings, thus eliminating the need for complicated exploratory behaviors including trajectory generation. By applying multiple pre-trained vision-language models to this omnidirectional image and incorporating reflective behaviors, we show that navigation becomes simple and does not require any prior setup. Interesting properties and limitations of our method are discussed based on experiments with the mobile robot Fetch.

Kento Kawaharazuka, Yoshiki Obinata, Naoaki Kanazawa et al.

In order for robots to autonomously navigate and operate in diverse environments, it is essential for them to recognize the state of their environment. On the other hand, the environmental state recognition has traditionally involved distinct methods tailored to each state to be recognized. In this study, we perform a unified environmental state recognition for robots through the spoken language with pre-trained large-scale vision-language models. We apply Visual Question Answering and Image-to-Text Retrieval, which are tasks of Vision-Language Models. We show that with our method, it is possible to recognize not only whether a room door is open/closed, but also whether a transparent door is open/closed and whether water is running in a sink, without training neural networks or manual programming. In addition, the recognition accuracy can be improved by selecting appropriate texts from the set of prepared texts based on black-box optimization. For each state recognition, only the text set and its weighting need to be changed, eliminating the need to prepare multiple different models and programs, and facilitating the management of source code and computer resource. We experimentally demonstrate the effectiveness of our method and apply it to the recognition behavior on a mobile robot, Fetch.

Kento Kawaharazuka, Kei Okada, Masayuki Inaba

Humans can autonomously learn the relationship between sensation and motion in their own bodies, estimate and control their own body states, and move while continuously adapting to the current environment. On the other hand, current robots control their bodies by learning the network structure described by humans from their experiences, making certain assumptions on the relationship between sensors and actuators. In addition, the network model does not adapt to changes in the robot's body, the tools that are grasped, or the environment, and there is no unified theory, not only for control but also for state estimation, anomaly detection, simulation, and so on. In this study, we propose a Generalized Multisensory Correlational Model (GeMuCo), in which the robot itself acquires a body schema describing the correlation between sensors and actuators from its own experience, including model structures such as network input/output. The robot adapts to the current environment by updating this body schema model online, estimates and controls its body state, and even performs anomaly detection and simulation. We demonstrate the effectiveness of this method by applying it to tool-use considering changes in grasping state for an axis-driven robot, to joint-muscle mapping learning for a musculoskeletal robot, and to full-body tool manipulation for a low-rigidity plastic-made humanoid.

Moju Zhao, Kei Okada, Masayuki Inaba

Template-based discriminative trackers are currently the dominant tracking methods due to their robustness and accuracy, and the Siamese-network-based methods that depend on cross-correlation operation between features extracted from template and search images show the state-of-the-art tracking performance. However, general cross-correlation operation can only obtain relationship between local patches in two feature maps. In this paper, we propose a novel tracker network based on a powerful attention mechanism called Transformer encoder-decoder architecture to gain global and rich contextual interdependencies. In this new architecture, features of the template image is processed by a self-attention module in the encoder part to learn strong context information, which is then sent to the decoder part to compute cross-attention with the search image features processed by another self-attention module. In addition, we design the classification and regression heads using the output of Transformer to localize target based on shape-agnostic anchor. We extensively evaluate our tracker TrTr, on VOT2018, VOT2019, OTB-100, UAV, NfS, TrackingNet, and LaSOT benchmarks and our method performs favorably against state-of-the-art algorithms. Training code and pretrained models are available at https://github.com/tongtybj/TrTr.

Kento Kawaharazuka, Naoaki Kanazawa, Yoshiki Obinata et al.

The state recognition of the environment and objects by robots is generally based on the judgement of the current state as a classification problem. On the other hand, state changes of food in cooking happen continuously and need to be captured not only at a certain time point but also continuously over time. In addition, the state changes of food are complex and cannot be easily described by manual programming. Therefore, we propose a method to recognize the continuous state changes of food for cooking robots through the spoken language using pre-trained large-scale vision-language models. By using models that can compute the similarity between images and texts continuously over time, we can capture the state changes of food while cooking. We also show that by adjusting the weighting of each text prompt based on fitting the similarity changes to a sigmoid function and then performing black-box optimization, more accurate and robust continuous state recognition can be achieved. We demonstrate the effectiveness and limitations of this method by performing the recognition of water boiling, butter melting, egg cooking, and onion stir-frying.

Hirokazu Ishida, Naoki Hiraoka, Kei Okada et al.

Library-based methods are known to be very effective for fast motion planning by adapting an experience retrieved from a precomputed library. This article presents CoverLib, a principled approach for constructing and utilizing such a library. CoverLib iteratively adds an experience-classifier-pair to the library, where each classifier corresponds to an adaptable region of the experience within the problem space. This iterative process is an active procedure, as it selects the next experience based on its ability to effectively cover the uncovered region. During the query phase, these classifiers are utilized to select an experience that is expected to be adaptable for a given problem. Experimental results demonstrate that CoverLib effectively mitigates the trade-off between plannability and speed observed in global (e.g. sampling-based) and local (e.g. optimization-based) methods. As a result, it achieves both fast planning and high success rates over the problem domain. Moreover, due to its adaptation-algorithm-agnostic nature, CoverLib seamlessly integrates with various adaptation methods, including nonlinear programming-based and sampling-based algorithms.

Kento Kawaharazuka, Yoshiki Obinata, Naoaki Kanazawa et al.

State recognition of the environment and objects, such as the open/closed state of doors and the on/off of lights, is indispensable for robots that perform daily life support and security tasks. Until now, state recognition methods have been based on training neural networks from manual annotations, preparing special sensors for the recognition, or manually programming to extract features from point clouds or raw images. In contrast, we propose a robotic state recognition method using a pre-trained vision-language model, which is capable of Image-to-Text Retrieval (ITR) tasks. We prepare several kinds of language prompts in advance, calculate the similarity between these prompts and the current image by ITR, and perform state recognition. By applying the optimal weighting to each prompt using black-box optimization, state recognition can be performed with higher accuracy. Experiments show that this theory enables a variety of state recognitions by simply preparing multiple prompts without retraining neural networks or manual programming. In addition, since only prompts and their weights need to be prepared for each recognizer, there is no need to prepare multiple models, which facilitates resource management. It is possible to recognize the open/closed state of transparent doors, the state of whether water is running or not from a faucet, and even the qualitative state of whether a kitchen is clean or not, which have been challenging so far, through language.

Kento Kawaharazuka, Kei Okada, Masayuki Inaba

In this study, we present an implementation strategy for a robot that performs peg transfer tasks in Fundamentals of Laparoscopic Surgery (FLS) via imitation learning, aimed at the development of an autonomous robot for laparoscopic surgery. Robotic laparoscopic surgery presents two main challenges: (1) the need to manipulate forceps using ports established on the body surface as fulcrums, and (2) difficulty in perceiving depth information when working with a monocular camera that displays its images on a monitor. Especially, regarding issue (2), most prior research has assumed the availability of depth images or models of a target to be operated on. Therefore, in this study, we achieve more accurate imitation learning with only monocular images by extracting motion constraints from one exemplary motion of skilled operators, collecting data based on these constraints, and conducting imitation learning based on the collected data. We implemented an overall system using two Franka Emika Panda Robot Arms and validated its effectiveness.

Moju Zhao, Tomoki Anzai, Kei Okada et al.

Two-dimensional multilinked structures can benefit aerial robots in both maneuvering and manipulation because of their deformation ability. However, certain types of singular forms must be avoided during deformation. Hence, an additional 1 Degrees-of-Freedom (DoF) vectorable propeller is employed in this work to overcome singular forms by properly changing the thrust direction. In this paper, we first extend modeling and control methods from our previous works for an under-actuated model whose thrust forces are not unidirectional. We then propose a planning method for the vectoring angles to solve the singularity by maximizing the controllability under arbitrary robot forms. Finally, we demonstrate the feasibility of the proposed methods by experiments where a quad-type model is used to perform trajectory tracking under challenging forms, such as a line-shape form, and the deformation passing these challenging forms.

Iori Yanokura, Naoki Wake, Kazuhiro Sasabuchi et al.

Learning actions from human demonstration video is promising for intelligent robotic systems. Extracting the exact section and re-observing the extracted video section in detail is important for imitating complex skills because human motions give valuable hints for robots. However, the general video understanding methods focus more on the understanding of the full frame,lacking consideration on extracting accurate sections and aligning them with the human's intent. We propose a Learning-from-Observation framework that splits and understands a video of a human demonstration with verbal instructions to extract accurate action sequences. The splitting is done based on local minimum points of the hand velocity, which align human daily-life actions with object-centered face contact transitions required for generating robot motion. Then, we extract a motion description on the split videos using video captioning techniques that are trained from our new daily-life action video dataset. Finally, we match the motion descriptions with the verbal instructions to understand the correct human intent and ignore the unintended actions inside the video. We evaluate the validity of hand velocity-based video splitting and demonstrate that it is effective. The experimental results on our new video captioning dataset focusing on daily-life human actions demonstrate the effectiveness of the proposed method. The source code, trained models, and the dataset will be made available.

Fan Shi, Timon Homberger, Joonho Lee et al.

Quadrupedal robots are skillful at locomotion tasks while lacking manipulation skills, not to mention dexterous manipulation abilities. Inspired by the animal behavior and the duality between multi-legged locomotion and multi-fingered manipulation, we showcase a circus ball challenge on a quadrupedal robot, ANYmal. We employ a model-free reinforcement learning approach to train a deep policy that enables the robot to balance and manipulate a light-weight ball robustly using its limbs without any contact measurement sensor. The policy is trained in the simulation, in which we randomize many physical properties with additive noise and inject random disturbance force during manipulation, and achieves zero-shot deployment on the real robot without any adjustment. In the hardware experiments, dynamic performance is achieved with a maximum rotation speed of 15 deg/s, and robust recovery is showcased under external poking. To our best knowledge, it is the first work that demonstrates the dexterous dynamic manipulation on a real quadrupedal robot.

Moju Zhao, Tomoki Anzai, Fan Shi et al.

Multilinked aerial robot is one of the state-of-the-art works in aerial robotics, which demonstrates the deformability benefiting both maneuvering and manipulation. However, the performance in outdoor physical world has not yet been evaluated because of the weakness in the controllability and the lack of the state estimation for autonomous flight. Thus we adopt tilting propellers to enhance the controllability. The related design, modeling and control method are developed in this work to enable the stable hovering and deformation. Furthermore, the state estimation which involves the time synchronization between sensors and the multilinked kinematics is also presented in this work to enable the fully autonomous flight in the outdoor environment. Various autonomous outdoor experiments, including the fast maneuvering for interception with target, object grasping for delivery, and blanket manipulation for firefighting are performed to evaluate the feasibility and versatility of the proposed robot platform. To the best of our knowledge, this is the first study for the multilinked aerial robot to achieve the fully autonomous flight and the manipulation task in outdoor environment. We also applied our platform in all challenges of the 2020 Mohammed Bin Zayed International Robotics Competition, and ranked third place in Challenge 1 and sixth place in Challenge 3 internationally, demonstrating the reliable flight performance in the fields.

Kentaro Wada, Kei Okada, Masayuki Inaba

We present joint learning of instance and semantic segmentation for visible and occluded region masks. Sharing the feature extractor with instance occlusion segmentation, we introduce semantic occlusion segmentation into the instance segmentation model. This joint learning fuses the instance- and image-level reasoning of the mask prediction on the different segmentation tasks, which was missing in the previous work of learning instance segmentation only (instance-only). In the experiments, we evaluated the proposed joint learning comparing the instance-only learning on the test dataset. We also applied the joint learning model to 2 different types of robotic pick-and-place tasks (random and target picking) and evaluated its effectiveness to achieve real-world robotic tasks.

Kentaro Wada, Shingo Kitagawa, Kei Okada et al.

We present a robotic system for picking a target from a pile of objects that is capable of finding and grasping the target object by removing obstacles in the appropriate order. The fundamental idea is to segment instances with both visible and occluded masks, which we call `instance occlusion segmentation'. To achieve this, we extend an existing instance segmentation model with a novel `relook' architecture, in which the model explicitly learns the inter-instance relationship. Also, by using image synthesis, we make the system capable of handling new objects without human annotations. The experimental results show the effectiveness of the relook architecture when compared with a conventional model and of the image synthesis when compared to a human-annotated dataset. We also demonstrate the capability of our system to achieve picking a target in a cluttered environment with a real robot.

Kentaro Wada, Kei Okada, Masayuki Inaba

Probabilistic 3D map has been applied to object segmentation with multiple camera viewpoints, however, conventional methods lack of real-time efficiency and functionality of multilabel object mapping. In this paper, we propose a method to generate three-dimensional map with multilabel occupancy in real-time. Extending our previous work in which only target label occupancy is mapped, we achieve multilabel object segmentation in a single looking around action. We evaluate our method by testing segmentation accuracy with 39 different objects, and applying it to a manipulation task of multiple objects in the experiments. Our mapping-based method outperforms the conventional projection-based method by 40 - 96\% relative (12.6 mean $IU_{3d}$), and robot successfully recognizes (86.9\%) and manipulates multiple objects (60.7\%) in an environment with heavy occlusions.

Kentaro Wada, Masaki Murooka, Kei Okada et al.

Picking objects in a narrow space such as shelf bins is an important task for humanoid to extract target object from environment. In those situations, however, there are many occlusions between the camera and objects, and this makes it difficult to segment the target object three dimensionally because of the lack of three dimentional sensor inputs. We address this problem with accumulating segmentation result with multiple camera angles, and generating voxel model of the target object. Our approach consists of two components: first is object probability prediction for input image with convolutional networks, and second is generating voxel grid map which is designed for object segmentation. We evaluated the method with the picking task experiment for target objects in narrow shelf bins. Our method generates dense 3D object segments even with occlusions, and the real robot successfuly picked target objects from the narrow space.