Direct Fusion Method Research Articles

Pseudo-Label Calibration Semi-supervised Multi-Modal Entity Alignment

Multi-modal entity alignment (MMEA) aims to identify equivalent entities between two multi-modal knowledge graphs for integration. Unfortunately, prior arts have attempted to improve the interaction and fusion of multi-modal information, which have overlooked the influence of modal-specific noise and the usage of labeled and unlabeled data in semi-supervised settings. In this work, we introduce a Pseudo-label Calibration Multi-modal Entity Alignment (PCMEA) in a semi-supervised way. Specifically, in order to generate holistic entity representations, we first devise various embedding modules and attention mechanisms to extract visual, structural, relational, and attribute features. Different from the prior direct fusion methods, we next propose to exploit mutual information maximization to filter the modal-specific noise and to augment modal-invariant commonality. Then, we combine pseudo-label calibration with momentum-based contrastive learning to make full use of the labeled and unlabeled data, which improves the quality of pseudo-label and pulls aligned entities closer. Finally, extensive experiments on two MMEA datasets demonstrate the effectiveness of our PCMEA, which yields state-of-the-art performance.

Comparative research on multi-source heterogeneous data fusion technologies

Both structured market data and unstructured financial news data can significantly affect stock price fluctuations. Therefore, relying only on a single data source for stock price trend prediction may produce information bias. This article combines these two types of heterogeneous data and systematically compares the effectiveness of different data fusion technologies to establish the optimal data integration strategy, thereby improving the accuracy of stock price prediction. Specifically, this paper innovatively proposes a stock price trend prediction model named MVL-SVM. The model successfully combines multi-perspective learning with support vector machine, and realizes the effective fusion of stock market data and financial news data. It is worth mentioning that the prediction accuracy of MVL-SVM model using direct fusion method is more than 30% higher than that of index modeling method, which indicates that direct fusion method can fully learn effective information from multiple information sources, thus significantly improving the prediction effect.

RFCT: Multimodal Sensing Enhances Grasping State Detection for Weak-Stiffness Targets

Accurate grasping state detection is critical to the dexterous operation of robots. Robots must use multiple modalities to perceive external information, similar to humans. The direct fusion method of visual and tactile sensing may not provide effective visual–tactile features for the grasping state detection network of the target. To address this issue, we present a novel visual–tactile fusion model (i.e., RFCT) and provide an incremental dimensional tensor product method for detecting grasping states of weak-stiffness targets. We investigate whether convolutional block attention mechanisms (CBAM) can enhance feature representations by selectively attending to salient visual and tactile cues while suppressing less important information and eliminating redundant information for the initial fusion. We conducted 2250 grasping experiments using 15 weak-stiffness targets. We used 12 targets for training and three for testing. When evaluated on untrained targets, our RFCT model achieved a precision of 82.89%, a recall rate of 82.07%, and an F1 score of 81.65%. We compared RFCT model performance with various combinations of Resnet50 + LSTM and C3D models commonly used in grasping state detection. The experimental results show that our RFCT model significantly outperforms these models. Our proposed method provides accurate grasping state detection and has the potential to provide robust support for robot grasping operations in real-world applications.

Comparison of deterministic and ensemble weather forecasts on ship sailing speed optimization

Ship sailing speed optimization models are constructed based on prediction of ship fuel consumption, whose accuracy is highly influenced by the quality of sea and weather information. In this study, we develop two fusion methods for combining external meteorological data with ship noon report data, including the rhumb line based fusion method and the direct fusion method, and compare them in terms of accuracy in providing meteorological data. Next, we propose a framework based on the better data fusion strategy for comparing the impacts of deterministic and ensemble weather forecasts on ship speed optimization performance, enabling the evaluation of ship fuel consumptions under different speed plans based on weather forecast data available before departure. Results show that speed optimization based on ensemble weather forecasts has greater potential than that based on deterministic weather forecasts to diminish ship fuel consumption and thus to reduce greenhouse gas emissions.

Gated Path Aggregation Feature Pyramid Network for Object Detection in Remote Sensing Images

Object detection in remote sensing images is a challenge because remote sensing targets have characteristics such as small geometries, an unfixed direction and multiple poses. Recent studies have shown that the accuracy of object detection can be improved using feature fusion. However, direct fusion methods regard each layer as being of equal importance and rarely consider the hierarchical structure of multiple convolutional layers, leading to redundancy and rejected information being rarely applied during the fusion process. To address these issues, we propose a gated path aggregate (GPA) network that integrates path enhancement and information filtering into an end-to-end integrated network. Specifically, we first quantitatively analyze the performance of different gating functions to select the most suitable gating function. Then, we explore the embedding of soft switchable atrous convolution (SSAC) in the topmost feature layer. Finally, we validate our proposed model by combining it with experiments using the public NWPU VHR-10 dataset. The experimental results show that our proposed GPAFPN structure has significant improvement compared to the FPN structure. Compared with the mainstream networks, it achieved state-of-the-art performance.

EL-SLE: Efficient Learning Based Stride-Length Estimation Using a Smartphone.

The pedestrian stride-length estimation is a crucial piece of personal behavior data for many smartphone applications, such as health monitoring and indoor location. The performance of the present stride-length algorithms is suitable for simple gaits and single scenes, but when applied to sophisticated gaits or heterogeneous devices, their inaccuracy varies dramatically. This paper proposes an efficient learning-based stride-length estimation model using a smartphone to obtain the correct stride length. The model uses adaptive learning to extract different elements for changing and recognition tasks, including Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN) modules. The direct fusion method maps the eigenvectors to the appropriate stride length after combining the features from the learning modules. We presented an online learning module to update the model to increase the SLE model’s generalization. Extensive experiments are conducted with heterogeneous devices or users, various gaits, and switched scenarios. The results confirm that the proposed method outperforms other state-of-the-art methods and achieves an average 4.26% estimation error rate in various environments.

Fusion method for sample preparation for isotopic composition determination of boron in refractory materials by thermal ionization mass spectrometry with validation using dissolved and purified samples

Various refractory samples of boron such as titanium borides, rare earth borides, boron carbides having natural and non-natural isotopic composition were analysed by thermal ionization mass spectrometry as Na2BO2+ ions. Samples were fused with Na2CO3 directly on the filament after adjusting the B/Na mole ratio to 2 and in certain cases modifications in the loading technique were required to obtain precise 10B/11B isotopic ratio. A few of the samples expected to have natural 10B/11B ratio were observed to be slightly enriched in 10B. Dissolved samples as well as boron purified by extraction with 5% 2-Ethyl Hexane Diol(EHD) were analysed for 10B/11B ratio to ascertain the accuracy of the results obtained by the direct fusion method. Precision of about 0.02% (1RSD) was obtained for 10B/11B isotopic ratio measured in all the three types of samples. A good agreement of the dissolved sample and purified boron fraction with the directly fused sample validated the direct fusion method as a fast and reliable technique for isotopic analysis of boron in refractory samples.

Building Damage Detection via Superpixel-Based Belief Fusion of Space-Borne SAR and Optical Images

Space-borne synthetic aperture radar (SAR) and optical sensors are important tools for building damage detection. Fusion of SAR and optical images improves detection performance. However, when the resolutions of the two different kinds of images differ, the performance of the existing pixel-level fusion methods deteriorates significantly due to interpolation-induced distortion. To solve this problem, this paper presents a new superpixel-based belief fusion (SBBF) model for building damage detection. The superpixels on the SAR and optical images are identified by the segmentation on the pre-earthquake optical image to perform the fusion on the superpixel-level instead of the pixel-level in existing methods. Then in the fusion stage, different from the commonly used direct fusion methods that do not consider the reliability in the fusion process, a novel belief fusion method that employs a basic belief assignment (BBA) to incorporate different reliabilities of superpixels is proposed to improve the accuracy of building damage detection. For each superpixel, the BBA is assigned based on the influence of noise and resolutions. The United Nations Operational Satellite Applications Programme (UNOSAT) datasets corresponding to the 2010 Haiti earthquake and the 2011 Tōhoku earthquake, are used to evaluate the performance of the proposed method. The experimental results show that the proposed method achieves significantly better performance than existing separate SAR or optical images based methods, and the existing pixel-level fusion methods.

Physical and chemical study of the dependence of substrate temperature on properties of Cu3SbS3 thin films

This work highlights the effect of substrate temperature on the properties of Cu3SbS3 thin films grown on unheated and heated glass substrates, by vacuum thermal evaporation. The as-deposited film is elaborated at room temperature (25 °C). The heated substrates are processed at various temperatures, ranging from 70 °C to 180 °C. Cu3SbS3 powder was successfully synthesized by direct fusion method. The obtained films were assayed for their structural, morphological, optical and electrical properties using several techniques, such as X-ray diffraction (XRD), scanning electron microscopy (SEM), Energy dispersive X-ray Spectrometry analysis (EDXS), UV‐VIS‐NIR spectrophotometer and hot probe method, respectively. First, the XRD analysis shows amelioration in the structural properties by increasing the substrate temperature. The microstructural parameters of all films were investigated such as crystallite size, strain and dislocation density. Second, the optical analysis shows that thin films have high absorption coefficients ranging between 105 and 106 cm−1. In addition, as per the models of Wemple–DiDomenico and Spitzer-Fan other optical and dielectric constants were estimated. Finally, the electrical study reveals that the p-type electrical conductivity is verified in all the samples.

A novel medical image fusion by combining TV-L1 decomposed textures based on adaptive weighting scheme

Medical image fusion involves combining multiple images of diverse modalities to acquire superior image quality while retaining the information of an individual image. The fusion process aids physicians to diagnose and assess the disease by increasing the visual information and clarity. Direct fusion methods often generate undesirable effects leading to distortion and poor contrast. To some extent, multi-scale decomposition (MSD) methods have achieved success in various image fusion problems. However, they suffer from ringing artifacts, due to the blur caused by strong edges in decomposition steps. Hence, to overcome these drawbacks, we propose an efficient and novel image fusion algorithm for constructing a fused image through total variation (TV-L1) model using an optimized adaptive weighting scheme. Our method focuses on transferring the prominent textural details of Magnetic Resonance Imaging (MRI) and visual details of Positron Emission Tomography (PET) by cartoon and texture decomposition. It consists of data fidelity term to force the cartoon component to remain close to original image and a regularization term to transfer gradients from original image to the fused image. The present work preserves structural and functional details with high contrast. Experiments were conducted using MRI and PET images collected from standard Brain Atlas Datasets. The qualitative and quantitative analysis of the work suggest that the proposed method achieves better quality and accuracy as compared with the available state-of-the-art techniques.

A Hierarchical Deep Fusion Framework for Egocentric Activity Recognition Using a Wearable Hybrid Sensor System.

Recently, egocentric activity recognition has attracted considerable attention in the pattern recognition and artificial intelligence communities because of its wide applicability in medical care, smart homes, and security monitoring. In this study, we developed and implemented a deep-learning-based hierarchical fusion framework for the recognition of egocentric activities of daily living (ADLs) in a wearable hybrid sensor system comprising motion sensors and cameras. Long short-term memory (LSTM) and a convolutional neural network are used to perform egocentric ADL recognition based on motion sensor data and photo streaming in different layers, respectively. The motion sensor data are used solely for activity classification according to motion state, while the photo stream is used for further specific activity recognition in the motion state groups. Thus, both motion sensor data and photo stream work in their most suitable classification mode to significantly reduce the negative influence of sensor differences on the fusion results. Experimental results show that the proposed method not only is more accurate than the existing direct fusion method (by up to 6%) but also avoids the time-consuming computation of optical flow in the existing method, which makes the proposed algorithm less complex and more suitable for practical application.

Phonon properties of ZnSnSb2 + Mn semiconductors: Raman spectroscopy

AbstractDoping of II–IV–V2 semiconductors opens up new opportunities for wide application. Addition of Mn to these materials induces formation of magnetic clusters, which are responsible for high‐temperature ferromagnetism. Our aim was to examine how the addition of Mn influences the optical and structural properties of ZnSnSb2 by micro‐Raman spectroscopy. For four samples of Zn1 − xMnxSnSb2 synthesized using the direct fusion method, with x = 0.027, 0.066, 0.076, and 0.086, Raman spectra were measured at room temperature in spectral range from 60 to 300 cm−1. The obtained results indicate that these are multiphase materials. Based on the size and shape of complex microstructures, which consist of different phases and clusters, dispersive and duplex or triplex types of microstructures can be identified. Existence of ZnSb, SnSb, and MnSb phases was confirmed. By analyzing the Raman spectra, phonons of ZnSb and SnSb are determined and they are consistent with the data from literature. Phonon properties of ZnSnSb2, as well as of MnSb, are experimentally obtained for the first time. On the basis of a shift of the ZnSnSb2 phonons, we found that some amount of Mn entered lattice and form Zn1 − xMnxSnSb2. Microstructures affect the physical properties and behavior of a material. Analysis of this complex semiconductors and obtained results are important for their optimization and customization for possible applications.

Effects of sulfurization on the optical properties of Cu2ZnxFe1-xSnS4 thin films

In order to prepare thin films of novel semiconductor materials that contain only earth abundant, low cost and nontoxic elements, Cu2ZnxFe1-xSnS4 ingots were successfully synthesized by direct fusion method. Crushed powders of these ingots were used as raw materials for the thermal evaporation. Cu2ZnxFe1-xSnS4 (with x = 0, 0.25, 0.5, 0.75 and 1) thin films were deposited on non-heated glass substrates by vacuum evaporation method. The as deposited films were sulfurized for 30 min at sulfurization temperature Ts = 400 °C. The effects of the sulfurization on the structural and optical properties of CZFTS films were realized by X-ray diffraction (XRD) and UV–Vis spectroscopy. XRD patterns show that all sulfurized CZFTS films were polycrystalline in nature with a preferential orientation along the (112) plane. CFTS films exhibit a stannite structure while CZTS films had a kesterite structure. Optical measurements showed that CZFTS films sulfurized at 400 °C exhibited an optical transmittance between 60 and 80% and all materials had relatively high absorption coefficients in the range of 104–105 cm−1. The band gap energies of sulfurized CZFTS films decreased from 1.71 to 1.50 eV with the increase of the Zn content. The dispersion of the refractive index was discussed in terms of the single oscillator model proposed by Wemple and DiDomenico and the optical parameters such as refractive index, extinction coefficient, oscillator energy and dispersion energy were calculated. The electrical free carrier susceptibility and the carrier concentration on the effective mass ratio were evaluated according to the model of Spitzer and Fan. The hot probe analysis showed that all sulfurized CZFTS films are p-type conductivity.

IMAGE FUSION FOR MULTIFOCUS IMAGES USING SPEEDUP ROBUST FEATURES

The multi-focus image fusion technique has emerged as major topic in image processing in order to generate all focus images with increased depth of field from multi focus photographs. Image fusion is the process of combining relevant information from two or more images into a single image. The image registration technique includes the entropy theory. Speed up Robust Features (SURF), feature detector and Binary Robust Invariant Scalable Key points (BRISK) feature descriptor is used in feature matching process. An improved RANDOM Sample Consensus (RANSAC) algorithm is adopted to reject incorrect matches. The registered images are fused using stationary wavelet transform (SWT).The experimental results prove that the proposed algorithm achieves better performance for unregistered multiple multi-focus images and it especially robust to scale and rotation translation compared with traditional direct fusion method.

Structure, optical and transport measurements on bulk magnetic semiconductor In1−xMnxSe

The compounds In1−xMnxSe (x = 0.0, 0.025, 0.05, 0.1 and 0.15) have been prepared by direct fusion method. X-ray powder diffraction measurements have been performed by step scanning mode. All the prepared samples are single phase and have wurtzite hexagonal crystal structure with space group P63/mmc. The crystal structure has been refined by Rietveld method using FULLPROF program. It is found that, the unit cell volume is increases with increasing Mn content which indicates that Mn 2+ ions go to In 3+ ions in the InSe lattice. Also, from the refined values of the crystallite size, it was found that the crystallite size decreases with increasing the Mn content. The energy band gap is calculated from the optical measurements and is found to increase as the Mn content increases. The electrical conductivity and the activation energy decreases with increasing Mn concentration. The coercivities, Hc increases with increasing Mn content. The susceptibility increases with increasing Mn content (x).

Synthesis, characterization and pharmacological evaluation of pyrimidine analogues for the management of Alzheimer’s disease

Event Abstract Back to Event Synthesis, characterization and pharmacological evaluation of pyrimidine analogues for the management of Alzheimer’s disease Rakesh Yadav1*, Monika Maan1 and Divya Yadav1 1 Banasthali University, Pharmaceutical Chemistry, India Alzheimer’s disease (AD) is a highly complex neurodegenerative disorder, which is characterized by the loss of memory and manifests in the cholinergic branch of central nervous system. Cholinesterase inhibitors have been found to be efficacious in reducing the symptoms of early medium stage of AD. There are number of drugs, which have been reported such as donepezil, galantamine, tacrine for the treatment of Alzheimer's. The pyrimidine scaffold is an important pharmacological structural component in well-known anti-AD drugs. In the present study, we have designed, synthesized and evaluated a series of 2, 4-disubstituted pyrimidine derivatives (RY-114 to RY-123) for the treatment of neurodegenerative diseases. The synthesis of a key intermediate i.e. 4-(substituted aniline)-2-chloropyrimidine was performed using 2, 4-dichloropyrimidine (1) as a starting material. The condensation of 1 with various substituted aniline in the presence of DIPEA and ethanol gave the desired key intermediate. The titled compounds were synthesized by condensation of the key intermediate with various primary or secondary substituted amines using direct fusion method in microwave or reflux in the presence of DIPEA and n-butanol. The structures of newly synthesized compounds were characterized using various spectral techniques viz. FT-IR, NMR, elemental analysis etc. The biological evaluation (in vivo or in vitro) of the synthesized derivatives RY-114 to RY-123 showed gradient effect with respect to the standard drug (Donezepil). The outcome of the work may lead to be selective and medicinally active chemical entities for the management of Alzheimer’s disease. Keywords: Galantamine, Tacrine, Alzheimer’s disease, donepezil, Neurodegenerative disorders Conference: 14th Meeting of the Asian-Pacific Society for Neurochemistry, Kuala Lumpur, Malaysia, 27 Aug - 30 Aug, 2016. Presentation Type: YIC04: Young Investigator Colloquium 4 Topic: 14th Meeting of the Asian-Pacific Society for Neurochemistry Citation: Yadav R, Maan M and Yadav D (2016). Synthesis, characterization and pharmacological evaluation of pyrimidine analogues for the management of Alzheimer’s disease. Conference Abstract: 14th Meeting of the Asian-Pacific Society for Neurochemistry. doi: 10.3389/conf.fncel.2016.36.00065 Copyright: The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers. They are made available through the Frontiers publishing platform as a service to conference organizers and presenters. The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated. Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed. For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions. Received: 04 Aug 2016; Published Online: 11 Aug 2016. * Correspondence: Dr. Rakesh Yadav, Banasthali University, Pharmaceutical Chemistry, Banasthali, Rajasthan, India, rakesh_pu@yahoo.co.in Login Required This action requires you to be registered with Frontiers and logged in. To register or login click here. Abstract Info Abstract The Authors in Frontiers Rakesh Yadav Monika Maan Divya Yadav Google Rakesh Yadav Monika Maan Divya Yadav Google Scholar Rakesh Yadav Monika Maan Divya Yadav PubMed Rakesh Yadav Monika Maan Divya Yadav Related Article in Frontiers Google Scholar PubMed Abstract Close Back to top Javascript is disabled. Please enable Javascript in your browser settings in order to see all the content on this page.

An automatic image fusion algorithm for unregistered multiply multi-focus images

The multi-focus image fusion technique provides a promising way to extend the depth of defocused images by combining multiple images with diverse focuses into a single focused one. In this paper, we present a robust and automated algorithm for the fusion of unregistered multiply multi-focus images. The motivation of our method lies in the fact that the source images are assumed to be perfectly aligned in the majority of previous research. Actually, the assumption is difficult to achieve in many practical situations. Hence, image registration method for multi-focus images is talked in this paper. We choose a multi-focus image as reference one in the image registration process by entropy theory. Speeded Up Robust Features (SURF) feature detector with Binary Robust Invariant Scalable Keypoints (BRISK) feature descriptor is used in the feature matching process. An improved RANdom Sample Consensus (RANSAC) algorithm is adopted to reject incorrect matches. The registered images are fused using stationary wavelet transform (SWT) with sym5 wavelet basis. The experimental results prove that the proposed algorithm achieves better performance for unregistered multiply multi-focus images, and it is especially robust to scale and rotation translation compared with traditional direct fusion method.

Efficient Niementowski synthesis of novel 1,3,10,12-tetra-substituted-8H-pyrido[2',3':4,5]pyrimido[6,1-b]quinazolin-8-ones

Novel 1,3,10,12-tetrasubstituted-8H-pyrido(2',3':4,5)pyrimido(6,1-b)quinazolin-8-ones were synthesised by a Niementowski reaction involving condensation of substituted anthranilic acids with a 5,7-disubstituted-3H-pyrido(2,3-d)pyrimidin-4-ones. Microwave irradiation in polyphosphoric acid media was used in order to improve reactions where classical or direct fusion method was limited. The synthesis of title compounds highlights a comparative study of the classical, microwave technique and direct fusion methods.

A direct fusion method of preparing silicate rock glasses for energy-dispersive electron microprobe analysis

A method of preparation of homogeneous glass beads by direct fusion of bulk rock powders on an iridium-strip heater, and the quantitative analysis of these glasses using an electron microprobe fitted with an Si(Li) detector, are described. Major element concentrations obtained for international geochemical reference standards are of high precision, and agree well with recommended values obtained by other methods.