We present a strategy for designing and building very general robot manipulation systems using general-purpose task-and-motion planner with both engineered learned modules that estimate properties affordances of unknown objects. Such are closed-loop policies map from RGB images, depth joint encoder measurements to position commands. show this leads intelligent behaviors even without priori knowledge regarding the set objects, their geometries, affordances. how these can be flexibly composed...

In crowdsourcing markets, a requester announces task and calls for contribution from potential participants. With strategic participants, the needs to reward participants introduce incentives of participation. However, it is natural ask whether worth introducing if total payment eliciting too high. This paper addresses such fundamental concern by designing frugal mechanism with minimum used procure amount service contributions demanded. We design two mechanisms provide participation while...

Traditional manual and drone coating maintenance methods have problems such as low efficiency, high risk, uneven spraying, which are difficult to meet the operation needs of offshore wind farms. To address this challenge, study proposed an intelligent platform that integrates a tracked vacuum adsorption mobile chassis, multi-sensor fusion hierarchical control system. The robot system uses Jetson Nano main device (equipped with EfficientDet-D7x model), combined STM32F103ZET6 auxiliary unit...

Camouflaged objects often blend in with their surroundings, making the perception of a camouflaged object more complex procedure. However, most neural-network-based methods that simulate visual information processing pathway creatures only roughly define general process, which deficiently reproduces process identifying objects. How to make modeled neural networks perceive as effectively is significant topic deserves further consideration. After meticulous analysis biological processing, we...

Security is a critical concern in low-power IoT (Internet of Things) wireless sensors because these resource constrained devices are easy to attack and meanwhile the energy constraint will consume lot run algorithms for security purposes. We study efficiency data transmission problem that use capacitors harvest while considering cost running algorithms. Energy harvesting with has characteristic rate varies over time, it getting slower as capacitor gets more energy. This observation result...

Traditional Chinese Medicine (TCM) is the crystallization of medical heritage for thousands years and plays a huge role in human health. How to integrate TCM products with modern information intelligent production means efficiently obtain standardized medicine services very worthy research direction. In this paper, decocting service taken as object, combined artificial intelligence, blockchain other technologies, an system trusted collaboration whole process innovatively proposed designed....

This paper investigates the problem of maximizing utility for job scheduling where each consists multiple tasks, task has and also extra if all tasks that are completed. We provide a 2-approximation algorithm single-machine case multi-machine problem. Both algorithms include two steps. The first step employs Earliest Deadline First method to compute with only utility, it is proved obtains optimal result this sub-problem. second Dynamic Programming without derives result. An approximation can...

This paper studies the problem of maximizing energy utilization for data transmission in sensors with periodical wireless charging process while taking into account thermal effect. Two classes problems are analyzed: one is case that can only a limited period time, and other long enough time. Algorithms proposed to solve analysis these algorithms also provided. For first problem, three subproblems studied, and, general we give an algorithm derive performance bound ( 1 - 2 m ) O P T E compared...

Matching tabular data to a knowledge graph (KG) is critical for understanding the semantic column types, relationships, and entities of table. Existing matching approaches rely heavily on core columns that represent primary subject which other in table depend. However, discovering these before table’s semantics challenging. Most prior works use heuristic rules, such as leftmost column, discover single while an insightful discovery set accurately captures dependencies between often...

We introduce uncertainty-aware object instance segmentation (UncOS) and demonstrate its usefulness for embodied interactive segmentation. To deal with uncertainty in robot perception, we propose a method generating hypothesis distribution of obtain set region-factored hypotheses together confidence estimates by making multiple queries large pre-trained models. This process can produce results that achieve state-of-the-art performance on unseen problems. The output also serve as input to...

Generalization to novel object configurations and instances across diverse tasks environments is a critical challenge in robotics. Keypoint-based representations have been proven effective as succinct representation for capturing essential features, establishing reference frame action prediction, enabling data-efficient learning of robot skills. However, their manual design nature reliance on additional human labels limit scalability. In this paper, we propose KALM, framework that leverages...

Abstract The multi-directionality of polished surface topography (PST) is a key index to characterize the quality and integrity, its prediction before actual machining great significance toolpath planning, verification, optimization. However, most related works reported thus far focus mainly on material removal modeling but few multi-directionality. Hence, this paper proposes fast method precisely predict PST for pad-polishing process freeform surfaces. With method, pressure distribution...

During the COVID-19 period, Lecture Recording System is of great significance to remote and flexible education, key issue lecture recording automation detection teacher students. To achieve fast accurate detection, We applied YOLO algorithm adopted Eagleeye pruning method reduce amount parameters calculation model. also designed developed Intelligent based on hisi Hi3531DV200. Experiments data set collected from Internet shows that pruned model can 94.8% precision 31.25FPS speed Hi3531DV200,...

The available surface defect detection methods for disposable wooden spoons still involve screening with the naked eye. This method is not only inefficient but also accompanied by problems such as false and missed detection. Therefore, this paper proposes a based on an improved YOLOv5 network model (YOLOv5-TSPP). uses K-Means ++ algorithm to cluster target samples in data set obtain anchor frames that are more line different scales. Coordinate Attention module added backbone of improve...