Consider a robot tasked with tidying a desk with a meticulously constructed Lego sports car. A human may recognize that it is not appropriate to disassemble the sports car and put it away as part of the "tidying....
详细信息
ISBN:
(数字)9798350384574
ISBN:
(纸本)9798350384581
Consider a robot tasked with tidying a desk with a meticulously constructed Lego sports car. A human may recognize that it is not appropriate to disassemble the sports car and put it away as part of the "tidying." How can a robot reach that conclusion? Although large language models (LLMs) have recently been used to enable commonsense reasoning, grounding this reasoning in the real world has been challenging. To reason in the real world, robots must go beyond passively querying LLMs and actively gather information from the environment that is required to make the right decision. For instance, after detecting that there is an occluded car, the robot may need to actively perceive the car to know whether it is an advanced model car made out of Legos or a toy car built by a toddler. We propose an approach that leverages an LLM and vision language model (VLM) to help a robot actively perceive its environment to perform grounded commonsense reasoning. To evaluate our framework at scale, we release the MessySurfaces dataset which contains images of 70 real-world surfaces that need to be cleaned. We additionally illustrate our approach with a robot on 2 carefully designed surfaces. We find an average 12.9% improvement on the MessySurfaces benchmark and an average 15% improvement on the robot experiments over baselines that do not use active perception. The dataset, code, and videos of our approach can be found at https://***/grounded_commonsense_reasoning/.
Autonomous driving is one of the most anticipated technologies of modern times. However, autonomous driving typically relies on high-cost hardware or requires extensive training of today's popular neural network m...
详细信息
A finite-time formation control scheme is proposed for a team of nonholonomic mobile robots with input saturation under a directed communication graph. The leader’s information is estimated by a distributed finite-ti...
详细信息
ISBN:
(数字)9798350307535
ISBN:
(纸本)9798350307542
A finite-time formation control scheme is proposed for a team of nonholonomic mobile robots with input saturation under a directed communication graph. The leader’s information is estimated by a distributed finite-time observer. Based on the distributed finite-time observer, the adaptive control approach, the input-output feedback linearization technique, and an input saturation model, a finite-time formation tracking control scheme is proposed. Then, rigorous proof is provided to demonstrate that all follower robots can hold a desirable geometric formation in finite time while tracking a leader robot simultaneously. Ultimately, a numerical example is displayed to verify the effectiveness of the designed formation control protocol.
The manta ray with large and flat pectoral fins is a typical median-paired fin mode propulsion fish with high stability, manoeuvrability, and efficiency. In this paper, a novel manta ray inspired robot with flexible p...
详细信息
ISBN:
(数字)9798331509644
ISBN:
(纸本)9798331509651
The manta ray with large and flat pectoral fins is a typical median-paired fin mode propulsion fish with high stability, manoeuvrability, and efficiency. In this paper, a novel manta ray inspired robot with flexible pectoral fins is developed. The pectoral fins are directly and independently driven by the waterproof servos, which are connected to the rigid leading edge and drives the entire flexible PVC fin to oscillate in order to maintain amplitude, frequency, and phase stability. The segmented flexible thin plate models to replace continuous deformation of the pectoral fins are employed. Fluid simulations and vortex structure analyses were carried out via XFlow based on lattice Boltzmann method. Subsequently, a detailed analysis of the hydrodynamic simulation results are conducted. The results indicated that the robot exhibited hydrody-namic characteristics similar to the natural manta rays, with vortices attached to the leading and trailing edges of the fins during flapping. The chordwise flexibility had a significant impact on propulsive force generation, offering a theoretical foundation for optimizing propulsion structures of biomimetic underwater robot.
Advancements in artificial intelligence (AI) have transformed robotics by enabling systems to autonomously execute complex tasks with minimal human involvement. Traditional methods, however, often depend on costly har...
详细信息
Electric vehicles, or EVs, have drawn a lot of attention lately as an eco-friendly way to cut carbon emissions and lessen reliance on fossil fuels. This work uses a neural network model called Long Short-Term Memory (...
详细信息
ISBN:
(数字)9798331511166
ISBN:
(纸本)9798331511173
Electric vehicles, or EVs, have drawn a lot of attention lately as an eco-friendly way to cut carbon emissions and lessen reliance on fossil fuels. This work uses a neural network model called Long Short-Term Memory (LSTM), which is effective at processing sequential input, to analyze user attitudes toward electric vehicles. Opinion data were gathered from the social media network Twitter between March 1st and March 31st, 2023, using the keyword “electric vehicles”. Based on the obtained opinion data, the LSTM approach is utilized to categorize sentiments as positive or negative. Implementation of the random oversampling technique is important to be applied in increasing the accuracy of the $\mathbf{n}$ recall model for the specified class. The study's conclusions show that the LSTM model performs admirably when it comes to sentiment analysis using the test dataset. The model evaluation yields a precision of $\mathbf{99.65 \%}$ and an accuracy rating of $\mathbf{97.79 \%}$ . The modeling process is carried out with variance splitting data where the highest accuracy is obtained from a comparison of training testing data of $\mathbf{90 \%}$ : $\mathbf{10 \%}$ .
The objective of this paper goal is to propose models to predict the diagnosis of acute, chronic or terminal chronic renal failure using supervised machine learning methods. We used medical data from the Nephrology Se...
The objective of this paper goal is to propose models to predict the diagnosis of acute, chronic or terminal chronic renal failure using supervised machine learning methods. We used medical data from the Nephrology Service of the Aristide Le Dantec Hospital in Dakar to build prediction models with supervised learning methods: Multilayer Perceptron, Support Vector Machines, Random Forest and Extreme Gradient Boosting. The model built with Extreme Gradient Boosting proved to be the best according to the performance measures we used, namely precision, recall, F1-score and accuracy. Follow, in order, the models built with Random Forest, Logistic Regression, Support Vector Machines and Multilayer Perceptron.
Wireless sensor networks (WSNs) are indispensable for various applications, including environmental monitoring and factory automation. Nevertheless, the precision with which sensor nodes can be localized is crucial to...
详细信息
Radar sensors are emerging as solutions for perceiving surroundings and estimating ego-motion in extreme weather conditions. Unfortunately, radar measurements are noisy and suffer from mutual interference, which degra...
Radar sensors are emerging as solutions for perceiving surroundings and estimating ego-motion in extreme weather conditions. Unfortunately, radar measurements are noisy and suffer from mutual interference, which degrades the performance of feature extraction and matching, triggering imprecise matching pairs, which are referred to as outliers. To tackle the effect of outliers on radar odometry, $a$ novel outlier-robust method called ORORA is proposed, which is an abbreviation of Outlier-RObust RAdar odometry. To this end, a novel decoupling-based method is proposed, which consists of graduated non-convexity (GNC)-based rotation estimation and anisotropic component-wise translation estimation (A-COTE). Furthermore, our method leverages the anisotropic characteristics of radar measurements, each of whose uncertainty along the azimuthal direction is somewhat larger than that along the radial direction. As verified in the public dataset, it was demonstrated that our proposed method yields robust ego-motion estimation performance compared with other state-of-the-art methods. Our code is available at https://***/url-kaist/outlier-robust-radar-odometry.
Numerical optimization provides a computational tool widely used to control robotic systems subject to constraints during their motion. However, many of the methods used to solve these problems lack treatment of physi...
详细信息
暂无评论