Multi-fingered robotic hands have potential to enable robots to perform sophisticated manipulation tasks. However, teaching a robot to grasp objects with an anthropomorphic hand is an arduous problem due to the high dimensionality of state and action spaces. Deep Reinforcement Learning (DRL) offers techniques to design control policies for this kind of problems without explicit environment or hand modeling. However, state-of-the-art model-free algorithms have proven inefficient for learning such policies. The main problem is that the exploration of the environment is unfeasible for such high-dimensional problems, thus hampering the initial phases of policy optimization. One possibility to address this is to rely on off-line task demonstrations, but, oftentimes, this is too demanding in terms of time and computational resources. To address these problems, we propose the A Grasp Pose is All You Need (G-PAYN) method for the anthropomorphic hand of the iCub humanoid. We develop an approach to automatically collect task demonstrations to initialize the training of the policy. The proposed grasping pipeline starts from a grasp pose generated by an external algorithm, used to initiate the movement. Then a control policy (previously trained with the proposed G-PAYN) is used to reach and grab the object. We deployed the iCub into the MuJoCo simulator and use it to test our approach with objects from the YCB-Video dataset. Results show that G-PAYN outperforms current DRL techniques in the considered setting in terms of success rate and execution time with respect to the baselines. The code to reproduce the experiments is released together with the paper with an open source license.
相關內容
Despite significant improvements in robot capabilities, they are likely to fail in human-robot collaborative tasks due to high unpredictability in human environments and varying human expectations. In this work, we explore the role of explanation of failures by a robot in a human-robot collaborative task. We present a user study incorporating common failures in collaborative tasks with human assistance to resolve the failure. In the study, a robot and a human work together to fill a shelf with objects. Upon encountering a failure, the robot explains the failure and the resolution to overcome the failure, either through handovers or humans completing the task. The study is conducted using different levels of robotic explanation based on the failure action, failure cause, and action history, and different strategies in providing the explanation over the course of repeated interaction. Our results show that the success in resolving the failures is not only a function of the level of explanation but also the type of failures. Furthermore, while novice users rate the robot higher overall in terms of their satisfaction with the explanation, their satisfaction is not only a function of the robot's explanation level at a certain round but also the prior information they received from the robot.
As of today, robots exhibit impressive agility but also pose potential hazards to humans using/collaborating with them. Consequently, safety is considered the most paramount factor in human-robot interaction (HRI). This paper presents a multi-layered safety architecture, integrating both physical and cognitive aspects for effective HRI. We outline critical requirements for physical safety layers as service modules that can be arbitrarily queried. Further, we showcase an HRI scheme that addresses human factors and perceived safety as high-level constraints on a validated impact safety paradigm. The aim is to enable safety certification of human-friendly robots across various HRI scenarios.
3D occupancy prediction holds significant promise in the fields of robot perception and autonomous driving, which quantifies 3D scenes into grid cells with semantic labels. Recent works mainly utilize complete occupancy labels in 3D voxel space for supervision. However, the expensive annotation process and sometimes ambiguous labels have severely constrained the usability and scalability of 3D occupancy models. To address this, we present RenderOcc, a novel paradigm for training 3D occupancy models only using 2D labels. Specifically, we extract a NeRF-style 3D volume representation from multi-view images, and employ volume rendering techniques to establish 2D renderings, thus enabling direct 3D supervision from 2D semantics and depth labels. Additionally, we introduce an Auxiliary Ray method to tackle the issue of sparse viewpoints in autonomous driving scenarios, which leverages sequential frames to construct comprehensive 2D rendering for each object. To our best knowledge, RenderOcc is the first attempt to train multi-view 3D occupancy models only using 2D labels, reducing the dependence on costly 3D occupancy annotations. Extensive experiments demonstrate that RenderOcc achieves comparable performance to models fully supervised with 3D labels, underscoring the significance of this approach in real-world applications.
Deep Neural Networks (DNNs) have drawn attention because of their outstanding performance on various tasks. However, deploying full-fledged DNNs in resource-constrained devices (edge, mobile, IoT) is difficult due to their large size. To overcome the issue, various approaches are considered, like offloading part of the computation to the cloud for final inference (split computing) or performing the inference at an intermediary layer without passing through all layers (early exits). In this work, we propose combining both approaches by using early exits in split computing. In our approach, we decide up to what depth of DNNs computation to perform on the device (splitting layer) and whether a sample can exit from this layer or need to be offloaded. The decisions are based on a weighted combination of accuracy, computational, and communication costs. We develop an algorithm named SplitEE to learn an optimal policy. Since pre-trained DNNs are often deployed in new domains where the ground truths may be unavailable and samples arrive in a streaming fashion, SplitEE works in an online and unsupervised setup. We extensively perform experiments on five different datasets. SplitEE achieves a significant cost reduction ($>50\%$) with a slight drop in accuracy ($<2\%$) as compared to the case when all samples are inferred at the final layer. The anonymized source code is available at \url{//anonymous.4open.science/r/SplitEE_M-B989/README.md}.
Task planning for robotic cooking involves generating a sequence of actions for a robot to prepare a meal successfully. This paper introduces a novel task tree generation pipeline producing correct planning and efficient execution for cooking tasks. Our method first uses a large language model (LLM) to retrieve recipe instructions and then utilizes a fine-tuned GPT-3 to convert them into a task tree, capturing sequential and parallel dependencies among subtasks. The pipeline then mitigates the uncertainty and unreliable features of LLM outputs using task tree retrieval. We combine multiple LLM task tree outputs into a graph and perform a task tree retrieval to avoid questionable nodes and high-cost nodes to improve planning correctness and improve execution efficiency. Our evaluation results show its superior performance compared to previous works in task planning accuracy and efficiency.
Recent years have witnessed many successful trials in the robot learning field. For contact-rich robotic tasks, it is challenging to learn coordinated motor skills by reinforcement learning. Imitation learning solves this problem by using a mimic reward to encourage the robot to track a given reference trajectory. However, imitation learning is not so efficient and may constrain the learned motion. In this paper, we propose instruction learning, which is inspired by the human learning process and is highly efficient, flexible, and versatile for robot motion learning. Instead of using a reference signal in the reward, instruction learning applies a reference signal directly as a feedforward action, and it is combined with a feedback action learned by reinforcement learning to control the robot. Besides, we propose the action bounding technique and remove the mimic reward, which is shown to be crucial for efficient and flexible learning. We compare the performance of instruction learning with imitation learning, indicating that instruction learning can greatly speed up the training process and guarantee learning the desired motion correctly. The effectiveness of instruction learning is validated through a bunch of motion learning examples for a biped robot and a quadruped robot, where skills can be learned typically within several million steps. Besides, we also conduct sim-to-real transfer and online learning experiments on a real quadruped robot. Instruction learning has shown great merits and potential, making it a promising alternative for imitation learning.
In aerial combat, dogfighting poses intricate challenges that demand an understanding of both strategic maneuvers and the aerodynamics of agile fighter aircraft. In this paper, we introduce TempFuser, a novel long short-term temporal fusion transformer designed to learn tactical and agile flight maneuvers in aerial dogfights. Our approach employs two distinct LSTM-based input embeddings to encode long-term sparse and short-term dense state representations. By integrating these embeddings through a transformer encoder, our model captures the tactics and agility of fighter jets, enabling it to generate end-to-end flight commands that secure dominant positions and outmaneuver the opponent. After extensive training against various types of opponent aircraft in a high-fidelity flight simulator, our model successfully learns to perform complex fighter maneuvers, consistently outperforming several baseline models. Notably, our model exhibits human-like strategic maneuvers even when facing adversaries with superior specifications, all without relying on explicit prior knowledge. Moreover, it demonstrates robust pursuit performance in challenging supersonic and low-altitude environments. Demo videos are available at //sites.google.com/view/tempfuser.
This work explores the use of military simulations in predicting and evaluating the outcomes of potential scenarios. It highlights the evolution of military simulations and the increased capabilities that have arisen due to the advancement of artificial intelligence. Also, it discusses the various applications of military simulations, such as developing tactics and employment doctrines, training decision-makers, evaluating new acquisitions, and developing new technologies. The paper then focuses on the Brazilian Air Force's efforts to create its own simulation tool, the Aerospace Simulation Environment (Ambiente de Simula\c{c}\~ao Aeroespacial -- ASA in Portuguese), and how this cloud-based service called ASA Simulation as a Service (ASA-SimaaS) can provide greater autonomy and economy for the military force. The main contribution of this work is to present the ASA-SimaaS solution as a means of empowering digital transformation in defense scenarios, establishing a partnership network, and improving the military's simulation capabilities and competitiveness.
To succeed in a Big Data strategy, you have to arm yourself with a wide range of data skills and best practices. This strategy can result in an impressive asset that can streamline operational costs, reduce time to market, and enable the creation of new products. However, several Big Data challenges may take place in enterprises when it comes to moving initiatives of boardroom discussions to effective practices. From a broader perspective, we take on this paper two very important challenges, namely modeling, and management. The main context here is to highlight the importance of understanding data modeling and knowing how to process complex data while supporting the characteristics of each model.
Object pose estimation underwater allows an autonomous system to perform tracking and intervention tasks. Nonetheless, underwater target pose estimation is remarkably challenging due to, among many factors, limited visibility, light scattering, cluttered environments, and constantly varying water conditions. An approach is to employ sonar or laser sensing to acquire 3D data, however, the data is not clear and the sensors expensive. For this reason, the community has focused on extracting pose estimates from RGB input. In this work, we propose an approach that leverages 2D object detection to reliably compute 6D pose estimates in different underwater scenarios. We test our proposal with 4 objects with symmetrical shapes and poor texture spanning across 33,920 synthetic and 10 real scenes. All objects and scenes are made available in an open-source dataset that includes annotations for object detection and pose estimation. When benchmarking against similar end-to-end methodologies for 6D object pose estimation, our pipeline provides estimates that are 8% more accurate. We also demonstrate the real world usability of our pose estimation pipeline on an underwater robotic manipulator in a reaching task.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191