Search

The GaoFen (GF)-5-02 satellite is one of the new generations of hyperspectral observation satellites launched by China in 2021. The directional polarimetric camera (DPC) is an optical sensor onboard the GF-5-02 satellite. The precipitable water vapor (PWV) is a key detection parameter of DPC. However, the existing PWV data developed using DPC data have significant errors due to the lack of the timely calibration of the two bands (865, 910 nm) of DPC used for PWV retrieval. In order to acquire DPC PWV data with smaller errors, a calibration method is developed for these two bands. The method consists of two parts: (1) calibrate the 865 nm band of the DPC using the cross-calibration method, (2) calibrate the 910 nm band of the DPC according to the calibrated 865 nm band of the DPC. This method effectively addresses the issue of the absence of a calibration method for the water vapor absorption band (910 nm) of the DPC. Regardless of whether AERONET PWV data or SuomiNET PWV data are used as the reference data, the accuracy of the DPC PWV data developed using calibrated DPC data is significantly superior to that of the DPC PWV data retrieved using data before recalibration. This means that the calibration method for the NIR bands of the DPC can effectively enhance the quality of DPC PWV data. [ABSTRACT FROM AUTHOR]

Purpose: This paper aims to establish an efficient maintenance management system tailored for healthcare facilities, recognizing the crucial role of medical equipment in providing timely and precise patient care. Design/methodology/approach: The system is designed to function both as an information portal and a decision-support system. A knowledge-based approach is adopted centered on Semantic Web Technologies (SWTs), leveraging a customized ontology model for healthcare facilities' knowledge capitalization. Semantic Web Rule Language (SWRL) is integrated to address decision-support aspects, including equipment criticality assessment, maintenance strategies selection and contracting policies assignment. Additionally, Semantic Query-enhanced Web Rule Language (SQWRL) is incorporated to streamline the retrieval of decision-support outcomes and other useful information from the system's knowledge base. A real-life case study conducted at the University Hospital Center of Oran (Algeria) illustrates the applicability and effectiveness of the proposed approach. Findings: Case study results reveal that 40% of processed equipment is highly critical, 40% is of medium criticality, and 20% is of negligible criticality. The system demonstrates significant efficacy in determining optimal maintenance strategies and contracting policies for the equipment, leveraging combined knowledge and data-driven inference. Overall, SWTs showcases substantial potential in addressing maintenance management challenges within healthcare facilities. Originality/value: An innovative model for healthcare equipment maintenance management is introduced, incorporating ontology, SWRL and SQWRL, and providing efficient data integration, coordinated workflows and data-driven context-aware decisions, while maintaining optimal flexibility and cross-departmental interoperability, which gives it substantial potential for further development. [ABSTRACT FROM AUTHOR]

The problem of finding the most reliable routing strategy on urban transportation networks refers to determining the time-adaptive routing policy that maximizes the probability of on-time arrival at a destination given an arrival time threshold. The problem is defined on a stochastic and time-dependent network that captures real-world transportation systems' inherent uncertainty and dynamism. To solve this problem, we present a dynamic programming–based algorithm that benefits from a node-time pairs queue implementation. In addition to improving the computational running time in most cases, this implementation supports different queue disciplines, leading to different algorithmic approaches: label-correcting and label-setting methods. We prove the correctness of the algorithm and derive its worst case time complexity. We present computational experiments over real-world, large-scale transportation networks with up to ∼33,000 nodes, showing that the algorithm is a viable alternative to existing state-of-the-art methods. It can be four times faster for relatively tight arrival time thresholds and is competitive for looser ones. We also present experiments assessing the different queue disciplines used within the algorithm, the gains of the node–time pairs queue implementation, and comparing optimal strategies obtained from reliability and travel time objectives. [ABSTRACT FROM AUTHOR]

The hippocampal formation is crucial for learning and memory, with submodule CA3 thought to be the substrate of pattern completion. However, the underlying synaptic and computational mechanisms of this network are not well understood. Here, we perform circuit reconstruction of a CA3 module using three dimensional (3D) electron microscopy data and combine this with functional connectivity recordings and computational simulations to determine possible CA3 network mechanisms. Direct measurements of connectivity schemes with both physiological measurements and structural 3D EM revealed a high connectivity rate, multi-fold higher than previously assumed. Mathematical modelling indicated that such CA3 networks can robustly generate pattern completion and replay memory sequences. In conclusion, our data demonstrate that the connectivity scheme of the hippocampal submodule is well suited for efficient memory storage and retrieval. [ABSTRACT FROM AUTHOR]

Underground hydrogen storage (UHS) plays a critical role in ensuring the stability and security of the future clean energy supply. However, the efficiency and reliability of UHS technology depend heavily on the careful and criteria-driven selection of suitable storage sites. This study presents a hybrid multi-criteria decision-making framework integrating the Analytical Hierarchy Process (AHP) and Preference Ranking Organisation Method for Enrichment of Evaluations (PROMETHEE) to identify and select the best hydrogen storage sites among depleted gas reservoirs in the UK. To achieve this, a new set of site selection criteria is proposed in light of the technical and economic aspects of UHS, including location, reservoir rock quality and tectonic characteristics, maximum achievable hydrogen well deliverability rate, working gas capacity, cushion gas volume requirement, distance to future potential hydrogen clusters, and access to intermittent renewable energy sources (RESs). The framework is implemented to rank 71 reservoirs based on their potential and suitability for UHS. Firstly, the reservoirs are thoroughly evaluated for each proposed criterion and then the AHP-PROMETHEE technique is employed to prioritise the criteria and rank the storage sites. The study reveals that the total calculated working gas capacity based on single-well plateau withdrawal rates is around 881 TWh across all evaluated reservoirs. The maximum well deliverability rates for hydrogen withdrawal are found to vary considerably among the sites; however, 22 % are estimated to have deliverability rates exceeding 100 sm3/d, and 63 % are located within a distance of 100 km from a major hydrogen cluster. Moreover, 70 % have access to nearby RESs developments, with an estimated cumulative RESs capacity of approximately 181 GW. The results highlight the efficacy of the proposed multi-criteria site selection framework. The top five highest-ranked sites for UHS based on the evaluated criteria are the Cygnus, Hamilton, Saltfleetby, Corvette, and Hatfield Moors gas fields. The insights provided by this study can contribute to informed decision-making, sustainable development, and the overall progress of future UHS projects within the UK and globally. [Display omitted] • A novel multi-criteria site selection framework is proposed for UHS in gas fields. • Eight selection criteria are proposed to capture the most essential UHS technical and economic aspects. • The model integrates the AHP and PROMETHEE methods for decision-making. • 71 gas reservoirs in the UK are ranked based on their suitability for conversion to UHS assets. [ABSTRACT FROM AUTHOR]

In automated storage and retrieval systems (AS/RSs), the utilization of intelligent algorithms can reduce the makespan required to complete a series of input/output operations. This paper introduces a simulation optimization algorithm designed to minimize the makespan in a realistic AS/RS commonly found in the steel sector. This system includes weight and quality constraints for the selected items. Our hybrid approach combines discrete event simulation with biased-randomized heuristics. This combination enables us to efficiently address the complex time dependencies inherent in such dynamic scenarios. Simultaneously, it allows for intelligent decision making, resulting in feasible and high-quality solutions within seconds. A series of computational experiments illustrates the potential of our approach, which surpasses an alternative method based on traditional simulated annealing. [ABSTRACT FROM AUTHOR]

The primary focus of the warehouse relocation problem is to systematically and efficiently move items already stored in the warehouse to newly planned storage locations, which is a critical issue for the successful operation of Automated Storage and Retrieval Systems (AS/RS). There is limited research on the warehouse relocation problem for AS/RS with multiple loading and unloading stations. To address this issue, this research proposed an approach to rearrange the materials stored in a unit-load AS/RS with multiple loading stations to a target assignment. A mathematical model based on integer programming was built to determine the optimal sequence for relocating storage locations and arrange the sequence of movement. Since this is at least NP-hard to solve, a heuristic algorithm was designed to solve this problem in several segments, enabling its application in practical scenarios with larger data scales to enhance the operational efficiency of automated storage systems. Computational experiments were conducted using various problem sizes to assess the performance of the proposed algorithm. Additionally, this study applied the proposed method to plan storage relocation for a large-scale automated AS/RS, which is operational at a computer hardware manufacturer in Taiwan, as a means to verify the feasibility and effectiveness of the method proposed in this research. [ABSTRACT FROM AUTHOR]

Two-stage remote sensing image captioning (RSIC) methods have achieved promising results by incorporating additional pre-trained remote sensing tasks to extract supplementary information and improve caption quality. However, these methods face limitations in semantic comprehension, as pre-trained detectors/classifiers are constrained by predefined labels, leading to an oversight of the intricate and diverse details present in remote sensing images (RSIs). Additionally, the handling of auxiliary remote sensing tasks separately can introduce challenges in ensuring seamless integration and alignment with the captioning process. To address these problems, we propose a novel cross-modal retrieval and semantic refinement (CRSR) RSIC method. Specifically, we employ a cross-modal retrieval model to retrieve relevant sentences of each image. The words in these retrieved sentences are then considered as primary semantic information, providing valuable supplementary information for the captioning process. To further enhance the quality of the captions, we introduce a semantic refinement module that refines the primary semantic information, which helps to filter out misleading information and emphasize visually salient semantic information. A Transformer Mapper network is introduced to expand the representation of image features beyond the retrieved supplementary information with learnable queries. Both the refined semantic tokens and visual features are integrated and fed into a cross-modal decoder for caption generation. Through extensive experiments, we demonstrate the superiority of our CRSR method over existing state-of-the-art approaches on the RSICD, the UCM-Captions, and the Sydney-Captions datasets [ABSTRACT FROM AUTHOR]

An MIAPP-NALT system is an order picking system in which cases are picked at multiple in-the-aisle pick positions (MIAPP) and storage and retrieval operations are performed by a narrow aisle lift truck (NALT). In this paper, the operation of such a system involving three classes of stock keeping units with random demands for storage and retrieval operations is modeled as an M/G/1 queue, where "customers" are storage and retrieval requests, the "server" is the NALT, and retrieval requests have non-preemptive priority over storage requests. Our goal is to explore a methodology and solution method to obtain the optimal layout design of a class-based MIAPP-NALT system with stochastic demands and priority service. To this end, an operation time model of the system is developed and the first two moments for the operation time are derived. To overcome the challenge in finding the desired optimal layout, a near-optimal layout obtained via a heuristic approach is obtained at first and is improved afterwards. Based on the optimal layout, some valuable queueing results demonstrate the benefit of using a priority-based discipline. Moreover, some useful insights regarding the selection of dedicated versus random storage policies are obtained. [ABSTRACT FROM AUTHOR]

In real-time object tracking systems, it is essential to assign a unique identifier or label to each tracked object to distinguish it from other objects to reduce the number of ID switching (ID-Sw) and improve tracking quality. To perform the object labeling process, massive metadata must be researched to determine the appropriate label. Despite the simplicity of this process, it becomes very complex in large-scale video surveillance networks, where the metadata becomes "big metadata," which negatively affects the performance of the real-time surveillance system. In this regard, we propose a new automatic label retrieval solution for efficient real-time tracking using the ABCCF-tree index structure to avoid such problems. The main advantages of the ABCCF-tree are its ability to index massive metadata and its logarithmic search complexity, which implicitly reduces the search time. Additionally, the quality of its searches, which ensures a consistent labeling of the tracked objects. The experimental evaluation was conducted on a publicly available dataset generated by a multi-camera containing different crowd activities. The experimental results showed excellent and promising performance. [ABSTRACT FROM AUTHOR]

Memory formation is typically divided into phases associated with encoding, storage, consolidation, and retrieval. The neural determinants of these phases are thought to differ. This study first investigated the impact of the experience of novelty in rats incurred at a different time, before or after, the precise moment of memory encoding. Memory retention was enhanced. Optogenetic activation of the locus coeruleus mimicked this enhancement induced by novelty, both when given before and after the moment of encoding. Optogenetic activation of the locus coeruleus also induced a slow-onset potentiation of field potentials in area CA1 of the hippocampus evoked by CA3 stimulation. Despite the locus coeruleus being considered a primarily noradrenergic area, both effects of such stimulation were blocked by the dopamine D1/D5 receptor antagonist SCH 23390. These findings substantiate and enrich the evidence implicating the locus coeruleus in cellular aspects of memory consolidation in hippocampus. [ABSTRACT FROM AUTHOR]

Background: The Picture Archiving and Communication System (PACS) plays a critical role in modern healthcare by preserving and retrieving medical images. To the best of our knowledge, this study stands as the first of its kind conducted within the Egyptian healthcare landscape, aiming to evaluate the transformative effects of PACS in an oncology setting specifically in the Minia Oncology Center. This study highlights the primary objective of PACS which lies in enhancing the routing, retrieval, and display capabilities of medical images within radiology departments. The PACS acts as a centralized repository for imaging data and reports. The transition from film-based imaging to PACS-based digital systems signifies a paradigm shift in healthcare, leveraging technological advancements to enhance patient care, streamline operations, and optimize resource utilization. Methods: Minia Oncology Centers is a governmental center and it belongs to the Ministry of Health and Population at Upper Egypt. It provides a lot of medical, surgical, and diagnostic oncological services for the Minia governorate population. The study involved a comparative analysis of various metrics pre and post-PACS implementation. Descriptive statistics such as means, standard deviations, and percentages were calculated to assess changes in image retrieval times, report delivery intervals, and cost-related variables. Results: The results of this study vividly demonstrate the transformative impact of PACS implementation within the Minia Oncology Center's radiology department. The PACS significantly enhanced workflow efficiency, substantially reduced image-reporting delivery times for routine and emergency cases, and notably streamlined patient data management. The statistical assessments conducted, including paired t-tests and chi-square tests, underscored the significance and reliability of these observed improvements post-PACS implementation. This is translated into better patient care and marked cost reductions associated with film-based systems. Conclusion: This study provides compelling evidence of PACS's substantial impact on the Minia Oncology Center's radiology department. The system's implementation resulted in improved workflow efficiency, faster image retrieval, and enhanced collaboration among healthcare professionals. The PACS not only facilitated better patient care through improved diagnostic accuracy but also significantly reduced operational costs associated with traditional film-based imaging methods. [ABSTRACT FROM AUTHOR]

The automated warehouses currently in widespread use maintain highly efficient operations at the cost of consuming a large amount of energy. This study first considers the problem of operation optimization in multi-shuttle automated storage and retrieval systems (AS/RSs) to reduce both time and energy consumption. A time and energy bi-objective integer programming model is proposed to jointly optimize the storage/retrieval (S/R) location assignment and S/R scheduling for a multi-shuttle AS/RS. The adaptive variable neighbourhood search algorithm and Lagrangian relaxation algorithm are developed to tackle large-size instances. The results demonstrate that the bi-objective model is effective at finding the compromise solution considering travel time and energy consumption simultaneously. The decrease in travel time is obtained at the expense of increasing energy consumption and vice versa. The study provides valuable insights for warehouse managers to operate a time-efficient and energy-efficient automated warehouse. [ABSTRACT FROM AUTHOR]

In the healthcare sector, medical records contain sensitive information about patients, so guaranteeing the confidentiality and integrity of it is essential. To improve the security of it, blockchain technology is being utilized. The blockchain is a type of distributed ledger and it keeps data securely while also generating trust without the need of third party. It has data storage constraint and Merkle tree preserves data integrity but it is inefficient when searching transactions within it. Hence this paper describes InterPlanetary File System (IPFS) based storage and modified bloom tree data structure which is a hybridization of bloom filter and Merkle tree for efficient searching. To protect data privacy, initially it encrypts medical records using ciphertext policy‐attribute based encryption and then the data stored on IPFS returns a hash value. To diminish the false positive rate (FPR), the hash returned by IPFS is stored in two parts of the bloom filter. The first part stores the data by using "k" non‐cryptographic hash function and second part stores the transformed data with the same hash function. The bloom tree is created using Merkle proof for verification of medical record in blockchain. The experiments show that the proposed method reduces the FPR rate and searching complexity is O(log2). [ABSTRACT FROM AUTHOR]

Proofs of Data Reliability are cryptographic protocols that provide assurance to a user that a cloud storage system correctly stores her data and has provisioned sufficient redundancy to be able to guarantee reliable storage service. In this paper, we consider distributed cloud storage systems that make use of erasure codes to guarantee data reliability. We propose two novel Proof of Data Reliability schemes, namely POROS and PORTOS, that on the one hand guarantees the retrieval of the outsourced data in their entirety through the use of proofs of data possession and on the other hand ensures the actual storage of redundancy. As a result, POROS and PORTOS are compatible with the current cloud computing model where the cloud autonomously performs all maintenance operations without any interaction with the user. Moreover, POROS and PORTOS delegate the burden of generating the redundancy to the cloud as well. The security of both solutions is proved in the face of a rational adversary whereby the cheating cloud provider tries to gain storage savings without increasing its total operational cost. [ABSTRACT FROM AUTHOR]

The adaptability of human populations to changing environments is often attributed to the human capacity for social learning, innovation, and culture. In rapidly changing environments, it has been shown that maintaining high levels of cultural variation is beneficial because it allows for efficient adaptation. However, in many theoretical models, a high level of cultural variation also implies that a large amount of useless and perhaps detrimental information must be maintained and used, leading to lower population fitness in general. Here, we begin to investigate this often conflicting relationship between adaptation and cultural variation. We explicitly allow for the interplay between social learning and environmental variability, alongside the capacity for "memory," i.e., the storage, retrieval, and forgetting of information. Here, memory allows individuals to retain unexpressed cultural variation, which does not directly impact adaptation. We show that this capacity for memory facilitates the evolution of social learning across a broader range of circumstances than previously thought. Results from this analysis may help to establish whether and when memory should be incorporated into cultural evolutionary models focused on questions of adaptation. [ABSTRACT FROM AUTHOR]

Hashing-based cross-modal retrieval methods have become increasingly popular due to their advantages in storage and speed. While current methods have demonstrated impressive results, there are still several issues that have not been addressed. Specifically, many of these approaches assume that labels are perfectly assigned, despite the fact that in real-world scenarios, labels are often incomplete or partially missing. There are two reasons for this, as manual labeling can be a complex and time-consuming task, and annotators may only be interested in certain objects. As such, cross-modal retrieval with missing labels is a significant challenge that requires further attention. Moreover, the similarity between labels is frequently ignored, which is important for exploring the high-level semantics of labels. To address these limitations, we propose a novel method called Cross-Modal Hashing with Missing Labels (CMHML). Our method consists of several key components. First, we introduce Reliable Label Learning to preserve reliable information from the observed labels. Next, to infer the uncertain part of the predicted labels, we decompose the predicted labels into latent representations of labels and samples. The representation of samples is extracted from different modalities, which assists in inferring missing labels. We also propose Label Correlation Preservation to enhance the similarity between latent representations of labels. Hash codes are then learned from the representation of samples through Global Approximation Learning. We also construct a similarity matrix according to predicted labels and embed it into hash codes learning to explore the value of labels. Finally, we train linear classifiers to map original samples to a low-dimensional Hamming space. To evaluate the efficacy of CMHML, we conduct extensive experiments on four publicly available datasets. Our method is compared to other state-of-the-art methods, and the results demonstrate that our model performs competitively even when most labels are missing. [ABSTRACT FROM AUTHOR]

With increasing demand for efficient archive management, robots have been employed in paper-based archive management for large, unmanned archives. However, the reliability requirements of such systems are high due to their unmanned nature. To address this, this study proposes a paper archive access system with adaptive recognition for handling complex archive box access scenarios. The system comprises a vision component that employs the YOLOV5 algorithm to identify feature regions, sort and filter data, and to estimate the target center position, as well as a servo control component. This study proposes a servo-controlled robotic arm system with adaptive recognition for efficient paper-based archive management in unmanned archives. The vision part of the system employs the YOLOV5 algorithm to identify feature regions and to estimate the target center position, while the servo control part uses closed-loop control to adjust posture. The proposed feature region-based sorting and matching algorithm enhances accuracy and reduces the probability of shaking by 1.27% in restricted viewing scenarios. The system is a reliable and cost-effective solution for paper archive access in complex scenarios, and the integration of the proposed system with a lifting device enables the effective storage and retrieval of archive boxes of varying heights. However, further research is necessary to evaluate its scalability and generalizability. The experimental results demonstrate the effectiveness of the proposed adaptive box access system for unmanned archival storage. The system exhibits a higher storage success rate than existing commercial archival management robotic systems. The integration of the proposed system with a lifting device provides a promising solution for efficient archive management in unmanned archival storage. Future research should focus on evaluating the system's performance and scalability. [ABSTRACT FROM AUTHOR]

The main function of a library is the availability of information resources. To achieve this goal, it is necessary to correctly store and retrieve documents and information in libraries and information centers. This study mainly aimed to design a model of challenges and opportunities for storing and retrieving information in Iran public libraries from the librarians' perspectives. The practical research approach is based on the grounded theory in terms of the purpose. Using the grounded theory qualitatively, sampling was conducted by targeted (judgmental) and snowball (chain) techniques for the theoretical saturation of data. Interviews were done with 26 senior public library experts with experience and expertise in this field. Results of the analysis of the data obtained from the interviews show that management, policy-making, and fiscal mechanisms were the primary underlying conditions. Individual and organizational factors were influential under intermediate (intervening) conditions. The approaches of the model of challenges and opportunities for storing and retrieving information include creating an environment for the employees' growth and development and modifying the organizational structure in public libraries of Iran. The presently designed model consists of adequate dimensions and components, casual, underlying, and mediating conditions, and constituting elements, which show the central phenomenon, consequences, and strategies for implementing public policy. Finally, practical suggestions were presented for improving the storage and retrieval of information in the Iranian public libraries. [ABSTRACT FROM AUTHOR]

In automated storage and retrieval systems (ASRS), adopting wireless power transfer (WPT) technology is an available way to avoid the drawbacks of the contact power supply system for multiple stackers. When stackers are dropping goods, it is meaningful to recycle the power generated by the motors and feedback to the other stackers or the ground power supply for improving the overall efficiency of ASRS. This article proposes a bidirectional WPT (BD-WPT) system for multiple receivers. Based on the values of transmission power and the power flow directions of receivers, a general model of the BD-WPT system is established with identifying the system operation modes and their boundaries. To achieve flexible switching among different operation modes and adapt the required transmission power with maximum efficiency point tracking (MEPT), a control strategy of the BD-WPT system is proposed. In the 2-kW experimental prototype built, the power demands can be met in real time, and the power fed back by the stackers can be supplied to other loads or recycled by the ground power supply with good dynamic performance. Compared with the system without MEPT, the efficiency of the proposed system is improved by 26.7% at most, and the maximum efficiency achieves 96.4%. [ABSTRACT FROM AUTHOR]

In this article, we introduce the printcase command, which outputs data from a specific observation into an easy-to-read Microsoft Word or PDF document. printcase allows analysts to focus on a single observation within a dataset and view that observation in its entirety. The output displays fields in table format, with all variables identified by their corresponding labels and all responses identified by their corresponding value labels. We explain how printcase works, give examples of circumstances under which this type of table-based quasiquestionnaire would be useful, and provide code for printing single observations. [ABSTRACT FROM AUTHOR]

Power equipment operation and maintenance (O&M) requires plenty of domain knowledge to improve equipment security and power grid reliability. However, most knowledge is implicitly represented by the semantic text, which is hard to be comprehended by algorithms and restricts the intelligence level of equipment O&M. Therefore, this article proposes an event knowledge graph system to automatically extract event knowledge from O&M reports and represent the knowledge by an event knowledge graph (KG). First, a bidirectional long short‐term memory network (BiLSTM) combined with conditional random field (CRF) is employed to recognize the entities. Next, an attention‐based sequence‐to‐sequence model is proposed to detect multiple events and extract document‐level representation vectors. Then, a tree‐based table filling strategy is utilized to complete the artificially designed event tables. Finally, the proposed system renders the event tables into an event KG. Verified by the experiments, the proposed system outperforms the baselines with the F‐measure of 83.5% for event extraction and 94.2% for event detection. When dealing with coexisting events and scattering arguments, the F‐measure of the proposed system achieves 74.8% and 79.0%, respectively. Moreover, the proposed system has positive effects on intelligent retrieval, fault diagnosis, and condition assessment, verifying its potential value for equipment O&M. [ABSTRACT FROM AUTHOR]

Image retrieval usually faces scale-variance issues as the amount of image data is rapidly increasing, which calls for more accurate retrieval technology. Besides, existing methods usually treat pair-image similarity as a binary value which indicates whether two images share either at least one common label or none of shared labels. However, such similarity definition cannot truly describe the similarity ranking for different numbers of common labels when handling the multi-label image retrieval problem. In this paper, a Feature Disentangling and Reciprocal Learning (FDRL) method is introduced with label-guided similarity to solve the above multi-label image retrieval problem. Multi-scale features are first extracted by BNInception network and then disentangled to the corresponding high- and low-correlation features under the guidance of estimated global correlations. After that, the disentangled features are combined through a reciprocal learning approach to enhance the feature representation ability. Final hash codes are learned based on the global features derived from BNInception network and the combined features generated by reciprocal learning. The whole network is optimized by the proposed label-guided similarity loss function which aims to simultaneously preserve absolute similarity for hard image pairs and relative similarity for soft image pairs. Experimental results on three public benchmark datasets demonstrate that the proposed method outperforms current state-of-the-art techniques. The code is online here: 'https://github.com/Yong-DAI/FDRL'. [ABSTRACT FROM AUTHOR]

The integrated scheduling problem in automated storage and retrieval systems (AS/RS) and the hybrid flowshop is critical for the realization of lean logistics and just-in-time distribution in manufacturing systems. The bi-objective model that minimizes the operation time in AS/RS and the makespan in the hybrid flowshop is established to optimize the problem. A mixed algorithm, named GA-MBO algorithm, is proposed to solve the model, which combines the advantages of the strong global optimization ability of genetic algorithm (GA) and the strong local search ability of migratory birds optimization (MBO). To avoid useless solutions, different cross operations of storage and retrieval tasks are designed. Compared with three algorithms, including improved genetic algorithm, improved particle swam optimization, and a hybrid algorithm of GA and particle swam optimization, the experimental results showed that the GA-MBO algorithm improves the operation efficiency by 9.48%, 19.54%, and 5.12% and the algorithm robustness by 35.16%, 54.42%, and 39.38%, respectively, which further verified the effectiveness of the proposed algorithm. The comparative analysis of the bi-objective experimental results fully reflects the superiority of integrated scheduling optimization. [ABSTRACT FROM AUTHOR]

This paper proposes a method named Double-scale Similarity with Rich Features for Cross-modal Retrieval (DSRF) to handle the retrieval task between images and texts. The difficulties of cross-modal retrieval are manifested in how to establish a good similarity metric and obtain rich accurate semantics features. Most existing approaches map different modalities data into a common space by category labels and pair relations, which is insufficient to model the complex semantic relationships of multimodal data. A new similarity measurement method (Double-scale similarity) is proposed, in which the similarity of multimodal data does not depend on category labels only, but also on the objects involved. The retrieval result in the same category without identical objects will be punished appropriately, while the distance between the correct result and query is further close. Moreover, a semantics features extraction framework is designed to provide rich semantics features for the similarity metric. Multiple attention maps are created to focus on local features from different perspectives and obtain numerous semantics features. Distinguish from other works that accumulate multiple semantic representations for averaging, we use LSTM only with forgetting gate to eliminate the redundancy of repetitive information. Specifically, the forgetting factor is generated for each semantics features, and a larger forgetting factor coefficient removes the useless semantics information. We evaluate DSRF on two public benchmark, DSRF achieves competitive performance. [ABSTRACT FROM AUTHOR]

How to obtain good retrieval performance in the case of few-shot labeled samples is the current research focus of Person Re-Identification. To facilitate formal analysis, we formally put forward the concept of Pseudo-Supervised Learning (PSL) to represent a series of research works based on label generation under few-shot condition. Through extensive investigations, we find that the main problem that needs to be solved of PSL is how we can improve the quality of pseudo-label. To solve this problem, in this work, we proposed a simple yet effective Heterogeneous Pseudo-Supervised Learning (H-PSL) framework based on classical PSL to implement asynchronous match, which boosts the feature expression and then a better label prediction in the following. Specifically, a novel isomer is constructed as the feature extractor and is trained with a much larger amount of pseudo-supervised data, i.e. , samples with pseudo-labels. In this way, the isomer obtains advanced feature expression. We then deliberately implement a cross-level asynchronous match mechanism between model and pseudo-supervised data. As a result, the quality of pseudo-label is greatly improved and the feature expression performance also be optimized accordingly. In addition, to make better use of pseudo-supervised data, we also designed a knowledge fusion strategy to integrate the pseudo labels and their confidence which are easily obtained by the base model and isomer. Encouragingly, knowledge fusion strategy further removes the noise-labeled samples from candidate data. We conduct experiments on four popular datasets to fully verify the universality of the proposed method. The experimental results show that the proposed method improves the performance of all compared baseline works. [ABSTRACT FROM AUTHOR]

As an important part of the "air–ground" integrated water quality monitoring system, the inversion of water quality from unmanned airborne hyperspectral image has attracted more and more attention. Meanwhile, unmanned aerial vehicles (UAVs) have the characteristics of small size, flexibility and quick response, and can complete the task of water environment detection in a large area, thus avoiding the difficulty in obtaining satellite data and the limitation of single-point monitoring by ground stations. Most researchers use UAV for water quality monitoring, they take water samples back to library or directly use portable sensors for measurement while flying drones at the same time. Due to the UAV speed and route planning, the actual sampling time and the UAV passing time cannot be guaranteed to be completely synchronized, and there will be a difference of a few minutes. For water quality parameters such as chromaticity (chroma), chlorophyll-a (chl-a), chemical oxygen demand (COD), etc., the changes in a few minutes are small and negligible. However, for the turbidity, especially in flowing water body, this value of it will change within a certain range. This phenomenon will lead to noise error in the measured suspended matter or turbidity, which will affect the performance of regression model and retrieval accuracy. In this study, to solve the quality problem of label data in a flowing water body, an unmanned airborne hyperspectral water quality retrieval experiment was carried out in the Xiao River in Xi'an, China, which verified the rationality and effectiveness of label denoising analysis of different water quality parameters. To identify noisy label instances efficiently, we proposed an instance selection scheme. Furthermore, considering the limitation of the dataset samples and the characteristic of regression task, we build a 1DCNN model combining a self attention mechanism (SAM) and the network achieves the best retrieving performance on turbidity and chroma data. The experiment results show that, for flowing water body, the noisy-label instance selection method can improve retrieval performance slightly on the COD parameter, but improve greatly on turbidity and chroma data. [ABSTRACT FROM AUTHOR]

Because of the calibration requirements of large-area infrared remote-sensing equipment, a large-area blackbody radiation source with the area of 1.05 m × 1.05 m is developed at the National Institute of Metrology (NIM), China. The temperature range of the blackbody is from room temperature to 573 K. A multiple temperature control technique is employed for improving the temperature stability and uniformity of the blackbody. The effects of the heating zone and material thickness on the temperature uniformity are investigated by thermodynamic simulation, and then the uniformity and stability of the blackbody temperature are tested. Retrieval results illustrate that the temperature uniformity of the blackbody radiation source system is less than ± 0.65 K. The temperature stability of the blackbody at all temperature points is less than 0.1 K in 10 min. The expanded combined standard uncertainty of the blackbody is 0.97 K at 373 K, 1.13 K at 473 K, and 1.53 K at 573 K. [ABSTRACT FROM AUTHOR]

The autonomous vehicle storage and retrieval system (AVS/RS) significantly improves the responsiveness and throughput of the traditional automated storage and retrieval system (AS/RS) in regard to handling unit loads. The AVS/RS consists of multiple tiers connected to an elevator system and is equipped with at least two autonomous vehicles, that is, a shuttle and satellite. Other necessary equipment are the lifts and input/output buffer areas. This paper aims to present and apply an original hybrid analytical-simulative model for the design of a deep-lane and multisatellite AVS-RS by evaluating and controlling the system performance. This AVS-RS is equipped with multiple free and non-free satellites for each tier. As an original contribution, this study reviews the literature on AVS/RS according to the introduction of multiple features categorized into five homogeneous groups: (1) rack configuration, (2) vehicle kinematics and configuration, (3) dispatching rules, (4) modeling approach, and (5) validation. Two of the most critical issues in existing research studies are the random arrival time of storage and retrieval transactions and the random storage policy. The proposed modeling approach is data-driven and based on realistic assumptions, filling the gap between the literature and real applications. This hybrid model is applied to a case study of the beverage industry according to a what-if comparative and competitive multiscenario analysis. This data-driven assessment supports the decision-making process on the number of satellites for each tier, while simultaneously controlling the service and waiting times, system throughput, and vehicle utilization. The analysis based on the maximum system throughput estimation demonstrates that introducing more than two satellites does not increase the productivity of the system. [ABSTRACT FROM AUTHOR]

Before infrared hyperspectral data are used in satellite data assimilation systems or retrieval systems, the quantitative analysis of data deviation is necessary. Based on RTTOV's (Radiative Transfer for TOVS) simulation data of FY-3E/HIRAS-II (Hyperspectral Infrared Atmospheric Sounder) and the observation data of HIRAS-II, we counted the bias of observation minus simulation (OMB) during an on-orbit test; analyzed the characteristics and reasons for the bias from the perspective of the FOV (field of view), the scanning angle of the instrument, the day and night, and the target temperature change; and analyzed the stability of the radiometric calibration accuracy. We also combined the results of the MetOp-C/IASI (infrared atmospheric sounding interferometer), a similar high-precision instrument, with the bias of OMB to compare and evaluate the FY-3E/HIRAS-II radiometric calibration accuracy. In the end, we found that the mean OMB bias of the long-wave and medium-wave infrared bands is within ±2 K, and the bias standard deviation is better than 2 K; the bias of each FOV is consistent and the bias of most channels is better than 2 K. The OMB bias of each channel is consistent with the changes in the angle of the instrument. The bias trend of long-wave and medium-wave infrared channels is more consistent with the deviation of the day and night; the bias of the short-wave infrared channel at night is lower than in the daytime. When counting the bias as the target temperature changed, the results showed that there are no obvious temperature dependencies in the long-wave and medium-wave infrared channels. This reflects that the instrument's non-linear effect is well ordered. We further evaluated the stability of the radiometric calibration accuracy through statistics from the OMB standard deviation of each channel of FY-3E/HIRAS-II. Most channel accuracy stability values were better than 0.1 K. We calculated that IASI and HIRAS-II OMB have double differences, and the results show that the double difference in most channels is better than 1 K. It shows that the HIRAS-II and IASI observations are highly consistent. Through the statistics of the OMB bias during the on-orbit test period of FY-3E/HIRAS-II, we fully evaluated its radiometric calibration accuracy and laid the foundation for FY-3E/HIRAS-II data to be used in the retrieval application and assimilation system. [ABSTRACT FROM AUTHOR]

Cross-media retrieval (CMR) offers a flexible retrieval experience across multiple modalities. Existing CMR approaches are constrained by the assumption that the paired modalities are available in training, and they leverage the data of all modalities to obtain a common representation. However, as dealing with the data from new modality, the previous all modalities need to be re-trained, compromising the flexibility and practicality of CMR. In this paper, we propose an approach termed learning Modality-Agnostic Representation for Scalable cross-media retrieval (MARS), which allows each modality to be trained independently. To be specific, MARS treats the label information as a distinct modality, and introduces a label parsing module LabNet to generate semantic representation for correlating different modalities. Meanwhile, MARS constructs the modality-specific representation module DataNet to obtain the modality-shared representation and modality-exclusive representation equipped with unbiased semantic classification. Technically, for the first modality, we jointly train the LabNet and its DataNet to preserve the semantic similarity between the Label-derived representation and the modality-shared representation. For new modalities, MARS employs the well-learned LabNet to extract the representation in labels, and then such representation is served as the privilege to guide the associated DataNet training via the same objective. Furthermore, we assign the same classifier to the representation module of all modalities for better semantic alignment. With the above schema, the obtained modality-shared representation is considered to be modality-agnostic. Extensive experiments on several benchmark multi-modality datasets demonstrate that the proposed MARS achieves better results than existing methods. [ABSTRACT FROM AUTHOR]

Cross-modal hashing has received growing interest to facilitating efficient retrieval across large-scale multi-modal data, and existing methods still face three challenges: 1) Most offline learning works are unsuitable for processing and training with streaming multi-modal data. 2) Current online learning methods rarely consider the potential interdependency between the label categories. 3) Existing supervised methods often utilize pairwise label similarities or adopt relaxation scheme to learn hash codes, which, respectively, require much computation time or accumulate large quantization loss during the learning process. To alleviate these challenges, this paper presents an efficient Online Label Consistent Hashing (OLCH) approach for streaming cross-modal retrieval. The proposed approach first exploits the relative similarity of semantic labels and utilizes the multi-class classification to derive the common semantic vector. Then, an online semantic representation learning framework is adaptively designed to preserve the semantic similarity across different modalities, and a mini-batch online gradient descent approach associated with forward–backward splitting is developed to discriminatively optimize the hash functions. Accordingly, the hash codes are incrementally learned with high discriminative capability, while avoiding high computation complexity to process the streaming data. Extensive experiments highlight the superiority of the proposed approach and show its very competitive performance in comparison with the state-of-the-arts. • Present an efficient online label consistent hashing approach to benefit cross-modal retrieval for streaming data. • Develop an online semantic representation learning framework to preserve the semantic similarity. • Propose a mini-batch online gradient descent approach to incrementally optimize the hash functions. • Exploit a forward–backward splitting scheme to learn the discriminative hash codes. [ABSTRACT FROM AUTHOR]

Longline fishery surveys are major contributors to groundfish species abundance estimates. Catchability is often assumed constant, but bait is lost or consumed at inconsistent rates during and across gear sets. Accounting for hook availability and gear saturation can improve abundance estimates, but these data are challenging to collect. Our objectives were to determine whether video camera systems can be used to collect these data from the Northeast Fisheries Science Center's Gulf of Maine Bottom Longline Survey and to explore the associations between hook disposition percentages and environmental characteristics. Electronic monitoring camera systems were deployed on partner vessels during two fall and two spring seasons to capture video of gear retrieval. During post-survey review, individual hooks were recorded as one of the following hook dispositions: bait, empty hook, fish, and invertebrate. Percentages of these hook dispositions were characterized across categorical (season, day or night, and bottom substrata) and continuous (bottom temperature, bottom depth, and substrate rugosity) survey station variables. Nonmetric multidimensional scaling and logistic regression analyses revealed correlations between retained bait and low bottom temperatures, the spring season, shallow depths, rough bottom substrata, and high terrain rugosity. These characteristics were typical of shallow, inshore, and topographically variable survey strata, suggesting that catchability differs among habitat types. This information will be incorporated into catch rate models to develop standardized indices of abundance for groundfish stock assessments. This study is an example of the utility of electronic technologies to supplement on board data collection during cooperative fishery surveys. • We examined bait retention and hook disposition of retrieved longline survey gear. • Retained bait was correlated with springtime, rugged seafloor, and daytime hauls. • The effect of depth on bait retention was significant in the spring but not fall. • Catchability differs among habitat types and with seasonal ecological interactions. • Hook disposition data will be used to improve catch estimates of groundfish species. [ABSTRACT FROM AUTHOR]

In order to train an end‐to‐end model for better performance in distinguishing superclasses and make the model more sensitive to inter‐class differences, the authors propose a Feature Context Network (FCNet), a novel ensemble framework for image retrieval based on clustering (EFC) and metric learning. In the authors' approach, the EFC framework consists of one common feature extractor and multiple learning branches. This tree‐like structure not only promotes the stability of the training process but also keeps the diversity by using multiple branches. Each branch is learnt by an independent random re‐labelling operation that distributes original classes into superclasses. In such a way, the framework thus can accelerate the training process. To make the EFC framework focus on the different attributes across the classes, the authors utilize k‐means clustering to divide the training set into subsets and train the model sequentially with those subsets. Meanwhile, to reduce the impact of over‐fitting problems, the authors design a branch weight structure during training. Further, FCNet fusions different scales of the neighbourhood information and consequently makes the feature map learn richer information with the different receptive fields. Simultaneously, it can reduce the layer of the network and the number of model parameters. Extensive experiments demonstrate that this approach outperforms the tested representative methods on CARS‐196, CUB‐200‐2011 datasets. [ABSTRACT FROM AUTHOR]

This is a report on the No.2 debris retrieval robot that has been developed. While unit 1 was a machine built to demonstrate basic ideas, such as the arm mechanism, the robot that was designed and manufactured this time was developed with an emphasis on its structure and the functions that can be used at the decommissioning site. Only the ideas of a crawler that can run through a thin pipe and a telescopic arm mechanism to access debris were carried over from unit 1; all other parts for debris retrieval robot No.2 were redesigned. The crawler was made from metal with significantly increased rigidity to improve the running performance of the robot. In addition, to improve maneuverability, an omnidirectional camera unit was installed. As the view of this camera is not limited to the direction of travel of the robot, it makes it possible for the operator to have a better grasp of situations that the robot may encounter. The debris collection mechanism was changed to a more reliable mechanism that is not affected by the shape of the debris. The arm expansion and contraction mechanisms were also changed to a more reliable structure. [ABSTRACT FROM AUTHOR]

Deep hashing has been widely used for large-scale cross-modal retrieval benefited from the low storage cost and fast search speed. However, most existing deep supervised methods only preserve the instance-pairwise relationship supervised by the semantic similarity matrix, which always inufficient heterogeneous correlation. Thus, we propose the Deep Discrete Cross-Modal Hashing with Multiple Supervision ( DDCH ms ) to further enhance the semantic consistency of heterogeneous modalities. It improves the performance of semantic information retrieval with the joint supervision of instance-pairwise, instance-labeled and class-wise similarities. Specifically, we firstly utilize the instance-pairwise similarity matrix to supervise the learning process of heterogeneous networks and it keeps the pairwise correlation from the perspective of instance-instance. Specially, we design a semantic network to fully exploit the semantic information implicated in labels, which is also used to supervise multi-modal networks on instance-label level. Furthermore, we propose the class-wise hash codes to cooperate with the intrinsic label matrix as the prototypes, and it guides the hash learning and further ensures the precision and compactness of the learned hash codes. In addition, we design different discrete optimization strategies to optimize the class-wise hash codes and unified hash codes, respectively. That avoids the optimization errors and ensures the high-quality of learned hash codes. Experiments on three popular datasets indicate that our method outperforms other state-of-the-art methods in terms of cross-modal retrieval. [ABSTRACT FROM AUTHOR]

With the explosive growth of multimodal data, cross-modal retrieval has drawn increasing research interests. Hashing-based methods have made great advancements in cross-modal retrieval due to the benefits of low storage cost and fast query speed. However, there still exists a crucial challenge to improve the accuracy of cross-modal retrieval due to the heterogeneity gap between modalities. To further tackle this problem, in this paper, we propose a new two-staged cross-modal retrieval method, called Deep Semantic Hashing with Dual Attention (DSHDA). In the first stage of DSHDA, a Semantic Label Network (SeLabNet) is designed to extract label semantic features and hash codes by training the multi-label annotations, which can make the learning of different modalities in a common semantic space and bridge the modality gap effectively. In the second stage of DSHDA, we propose a deep neural network to simultaneously integrate feature and hash code learning for each modality into the same framework, the training of the framework is guided by the label semantic features and hash codes generated from SeLabNet to maximize the cross-modal semantic relevance. Moreover, dual attention mechanisms are used in our neural networks: (1) Lo-attention is used to extract the local key information of each modality and improve the quality of modality features. (2) Co-attention is used to strengthen the relationship between different modalities to produce more consistent and accurate hash codes. Extensive experiments on two real datasets with image-text modalities demonstrate the superiority of the proposed method in cross-modal retrieval tasks. [ABSTRACT FROM AUTHOR]

[Display omitted] • Proposed a novel two-step label embedding semantic-guided hashing method. • Proposed a fast alternative optimization strategy to address the non-convex problem. • Evaluated the effectiveness of the proposed method via three famous datasets. Hashing technologies have been widely used for information retrieval tasks due to their efficient retrieval and storage capabilities. Generally, most of the current supervised learning only utilizes labels to construct a binary similarity matrix of instance pairs and ignores the rich semantic information contained in the labels. Indeed, the reason why supervised hashing is better than unsupervised hashing is that the labels itself has strong discriminative information. Therefore, how to effectively explore the label information is one of the ways to improve the performance of retrieval tasks. In addition, existing hashing methods have the problems of high time consumption and weak scalability when facing large-scale data. To remedy these problems, in this paper, we present a flexible two-step label embedding hashing method named L abel E mbedding S emantic- G uided H ashing (LESGH). In the first step, LESGH leverages an asymmetric discrete learning framework to learn discriminative compact hash codes only from label information, and adds the constraints of bit-balance and bit-decorrelation to boost the quality of the hash code generation. In the second step, LESGH learns the hash projection function through the generated hash codes in the first step. Moreover, an effective and fast iterative discrete optimization algorithm is presented to solve the discrete problem instead of using the relaxation-based scheme. In doing so, we can not only simplify the optimization process, but also easily scale to large-scale data. We conduct several experiments on three public datasets, i.e., WIKI, MIRFlickr and NUS-WIDE, demonstrate that LESGH can improve the retrieval performance over the compared state-of-the-art baselines. [ABSTRACT FROM AUTHOR]

Video moment retrieval has progressed significantly over the past few years, aiming to search the moment that is most relevant to a given natural language query. Most existing methods are trained in a fully-supervised or a weakly-supervised manner, which requires a time-consuming and expensive manually labeling process. In this work, we propose an alternative approach to achieving video moment retrieval that requires no textual annotations of videos and instead leverages the existing visual concept detectors and a pre-trained image-sentence embedding space. Specifically, we design a video-conditioned sentence generator to produce a suitable sentence representation by utilizing the mined visual concepts in videos. We then design a GNN-based relation-aware moment localizer to reasonably select a portion of video clips under the guidance of the generated sentence. Finally, the pre-trained image-sentence embedding space is adopted to evaluate the matching scores between the generated sentence and moment representations with the knowledge transferred from the image domain. By maximizing these scores, the sentence generator and moment localizer can enhance and complement each other to achieve the moment retrieval task. Experimental results on the Charades-STA and ActivityNet Captions datasets demonstrate the effectiveness of our proposed method. [ABSTRACT FROM AUTHOR]

Extracting high resolution, shallow bathymetry from drone-based imagery using empirical methods is a novel approach that gains increasing interest. However, due to the rapidly expanding character of this field, there are not sufficient guidelines for optimal pre-processing of drone-based imagery regarding bathymetry retrieval. This study explores the suitability of imagery resulted from commonly used drone cameras in terms of producing good correlation with ground-truth depth measurements obtained from an unmanned surface vehicle (USV). Thus, imagery with similar spectral responses, from two commercial sensors (DJI Phantom 4 Pro and MicaSense-RedEdge multispectral camera), is examined along with and without radiometric corrections applied, using proprietary software. The results show that radiometric image corrections using a reference reflectance panel, greatly enhance the correlation coefficient between bathymetry data and single-band or band-ratio reflectance mosaics at two study areas with mixed seafloor types. However, it was observed that the Red band and its associated logarithmic band ratios (Blue/Red, Green/Red) from both sensors show good correlation with depth as well, regardless of radiometric corrections. This result applies mainly to areas with favourable optical properties of water. This study suggests that radiometric calibration is crucial in empirical bathymetry retrieval from drone-based imagery. Furthermore, radiometric corrections applied on common Red, Green, Blue (RGB) cameras provide useful reflectance mosaics at lower costs compared to multispectral cameras. [ABSTRACT FROM AUTHOR]

This paper studies multiple-deep automated vehicle storage and retrieval systems (AVS/RSs) known for their high throughput performance and flexibility. Compared to a single-deep system, multiple-deep AVS/RS has a better space area utilisation. However, a relocation cycle occurs, reducing the throughput performance whenever another stock keeping unit (SKU) blocks a retrieving SKU. The SKU retrieval sequence is undetermined, meaning that the arrangement is unknown, and all SKUs have an equal probability of retrieval. In addition to the shuttle carrier, a satellite vehicle is attached to the shuttle carrier and is used to access storage locations in multiple depths. A discrete event simulation of multiple-deep AVS/RS with a tier captive shuttle carrier was developed. We focused on the dual-command cycle time assessment of nine different storage and relocation assignment strategy combinations in the simulation model. The results of a simulation study for (i) random, (ii) depth-first and (iii) nearest neighbour storage and relocation assignment strategy combinations are examined and benchmarked for five different AVS/RS case study configurations with the same number of storage locations. The results display that the fivefold- and sixfold-deep AVS/RSs outperform systems with fewer depths by utilising depth-first storage and nearest neighbour relocation assignment strategies. [ABSTRACT FROM AUTHOR]

This research focuses on big experimental data analysis and the replenishment operation in a crane & shuttle based storage and retrieval system (C&SBS/RS). It is proposed that the order structure is expressed by three indicators: the order strength, the order size, and the order density. This aims to develop a regression function for the replenishment strategy to reveal the relationship between the order structure and replenishment parameters. The simulation model of the system is developed using FlexSim 7.1. A total of 13,440 combinations of order structures are provided as the simulation input. The statistical analyses are completed using the popular statistical software MINITAB. In this study, we use two different methods, best subset regression, and stepwise regression, to fit the regression function and find the best regression. According to the regression function, operation managers can develop better operation strategies. [ABSTRACT FROM AUTHOR]

Deep cross-modal hashing which integrates deep learning and hashing into cross-modal retrieval, achieves better performance than traditional cross-modal retrieval methods. Nevertheless, most previous deep cross-modal hashing methods only utilize single-class labels to compute the semantic affinity across modalities but overlook the existence of multiple category labels, which can capture the semantic affinity more accurately. Additionally, almost all existing cross-modal hashing methods straightforwardly employ all modalities to learn hash functions but neglect the fact that original instances in all modalities may contain noise. To avoid the above weaknesses, in this paper, a novel multi-label enhancement based self-supervised deep cross-modal hashing (MESDCH) approach is proposed. MESDCH first propose a multi-label semantic affinity preserving module, which uses ReLU transformation to unify the similarities of learned hash representations and the corresponding multi-label semantic affinity of original instances and defines a positive-constraint Kullback–Leibler loss function to preserve their similarity. Then this module is integrated into a self-supervised semantic generation module to further enhance the performance of deep cross-modal hashing. Extensive evaluation experiments on four well-known datasets demonstrate that the proposed MESDCH achieves state-of-the-art performance and outperforms several excellent baseline methods in the application of cross-modal hashing retrieval. Code is available at: https://github.com/SWU-CS-MediaLab/MESDCH. [ABSTRACT FROM AUTHOR]

This article deals with cycle time calculation of Automated Storage and Retrieval Systems (AS/RS). Cycle time has a high impact on the operating performance of an AS/RS, and its knowledge is essential, both at the operational and design level. The novelty of this work concerns the peculiar kind of system that is considered, as the focus is on the Shuttle-Lift-Crane AS/RS. This solution, common in the steel sector, is used to store bundles of long metal bars, which are automatically handled by cranes, lifts, and shuttles. The functioning of these machines, which can operate in parallel and independently, is stochastically modeled, and the probability distribution function of the cycle time is computed, both for single and dual command cycles. The model, assessed via discrete event simulation, ensures a high average accuracy of 96% and 98%, under single and dual command cycles, respectively. [ABSTRACT FROM AUTHOR]

The availability of large-scale remote sensing (RS) data facilitates a wide range of applications, such as disaster management and urban planning. An approach for such problems is image retrieval, where, given a query image, the goal is to find the most relevant match from a database. Most RS literature has been focused on single-label retrieval, where we assume an image has a single label. The primary challenge in single-label RS retrieval is that performance in most datasets is saturated, and it has become difficult to compare the performance of different methods. In this work, we extend the major multilabel classification datasets to the multilabel retrieval problem. We also define protocols, provide evaluation metrics, and study the impact of commonly used loss functions and reranking methods for multilabel retrieval. To this end, a novel multilabel loss function and a reranking technique are proposed, which circumvent the challenges present in conventional single-label image retrieval. The developed loss function considers both class and feature similarity. The proposed reranking technique achieves high performance with computation cost that is well-suited for fast online retrieval. [ABSTRACT FROM AUTHOR]

In industrial systems, textual failure records note the failure mechanisms, the parts involved, and the failure symptoms; these records guide fault analysis and repair. However, case retrieval and feature extraction require extensive prior knowledge and diagnostic expertise; this method is time-consuming and labor-intensive. In this article, we present a novel two-stage framework that automatically extracts and generates features from very large textual records. We use an improved weighted latent Dirichlet allocation model and the Word2vec method to extract topic category and semantic features from fault texts; this approach accelerates training convergence. Next, we build a topic-context attention model in which word-embedding semantic features interact with topic features. Finally, we use classification and similarity calculation models to diagnose faults and retrieve similar cases; this approach ensures feature generation. Our method is very granular in terms of case representation which significantly improves diagnosis. The method robustly identifies similar cases by interrogating vehicle maintenance records. [ABSTRACT FROM AUTHOR]