Search

Video anomaly detection is an important task for public security in the multimedia field. It aims to distinguish events that deviate from normal patterns. As important semantic representation, the textual information can effectively perceive different contents for anomaly detection. However, most existing methods primarily rely on visual modality, with limited incorporation of textual modality in anomaly detection. In this paper, a cross-modality integration framework (CIForAD) is proposed for anomaly detection, which combines both textual and visual modalities for prediction, perception and discrimination. Firstly, a feature fusion prediction (FUP) module is designed to predict the target regions by fusing the visual features and textual features for prompting, which can amplify the discriminative distance. Then an image-text semantic perception (ISP) module is developed to judge semantic consistency by associating the fine-grained visual features with textual features, where a strategy of local training and global inference is introduced to perceive local details and global semantic correlation. Finally, a self-supervised time attention discrimination (TAD) module is built to explore the inter-frame relation and further distinguish abnormal sequences from normal sequences. Extensive experiments on the three challenging benchmarks indicate that our CIForAD obtains state-of-the-art anomaly detection performance. • Cross-modality integration framework (CIForAD) for unsupervised anomaly detection. • Prompts for frame prediction by fusing visual and textual features. • Semantic correlation maintained by image-text semantic perception module. • Time attention strategy on inter-frame relation for distinguishing time differences. • Extensive evaluations show outstanding anomaly detection performance. [ABSTRACT FROM AUTHOR]

With the development of the micro-electronics products, micro-structure become a research focus for its’ plenty of advantages, for example, high specific surface area and the application of surface wettability to enhance heat transfer. PIII technology is an advanced technology for preparing micro-structure surface, but there is few research about the method of preparing specified micro-structure surface. The micro-structure surface is prepared with PIII technology and characterized by SEM and AFM. The numerical work is conducted to investigate the impact of waveform and the structure parameters on surface contact angle. The results demonstrate that different waveform of the surface make a significant difference in contact angle. What’s more, with the same waveform, different wavelength and amplitude make the contact angle change. Finally, the numerical results make a good consistent with the experimental data with the increase of area ratio. [ABSTRACT FROM AUTHOR]

Acoustic energy harvesters are promising platforms for noise-to-electricity conversion in distributed environmental power supply and sound sensing in intuitive voice user interface. However, their efficacy in real-world applications is frustrated by the low electric output. We here report a multi-hole acoustic resonator enhanced triboelectric nanogenerator (MHAR-TENG) that couples a perforated plate resonator and a pressure differential acoustic receiver. It generated open-circuit voltage of 347 V, short-circuit current of 95 µA, and record-high power density of 8.9 W/m2. The numerical simulation confirmed the sound pressure amplification for enhanced output performance. Besides, the high signal-to-noise ratio and precise high-frequency response of electrical signals enabled the high-quality restoration of music and human speech. Assisted by machine learning, the MHAR-TENG realized self-powered voiceprint identification and emotion recognition with a remarkably improved accuracy over 90 %. Our MHAR-TENG holds great potential for efficient acoustic-to-electrical conversion in the fields of noise recycling, biometric authentication, and human-machine interaction. [Display omitted] • An acoustic nanogenerator by coupling the perforated resonator and the pressure differential acoustic receiver is proposed. • A record-high power density of 8.9 W/m2 is achieved under a low frequency excitation of 150 Hz and 104 dB. • The MHAR-TENG provides a new model for acoustic-to-electrical conversion in noise recycling and intelligent voice sensing. [ABSTRACT FROM AUTHOR]

Abstract: The black multi-crystalline silicon (mc-Si) has been successfully produced by plasma immersion ion implantation. The microstructure and the reflectance of the black mc-Si have been investigated by atomic force microscope and spectrophotometer, respectively. Results show that the black mc-Si exhibits a hillock structure with a low reflectance. Besides, with decreasing the diffusion temperature, the external quantum efficiency of the black mc-Si solar cell increases below ∼550nm wavelength due to reduced surface recombination. The optimal conversion effieciency of the black mc-Si solar cell is 15.50% at the diffusion temperature of 825°C. Furthermore, it is interesting to find that there are something different between black mc-Si and acid etched mc-Si on the impact of diffusion. [Copyright &y& Elsevier]

• Propose Cycle Reinforce Hierarchical Model for recognition tasks. • Design hierarchical structure composed of several hidden layers for features evolution. • Adopt groups of fuzzy subsystems with fuzzy rules ensures the diversity of features. • Introduce cycle mechanism for transferring refined coefficients. • Measure innovative strategies by the perspective of contribution rate. Neuro-fuzzy models have been applied in various domains, in which the issue of long time-consumption for optimizing parameters and less innovation in fuzzy method for feature extraction remains to be solved. Here, we present a novel cycle reinforce hierarchical model (CRHM) for effective and efficient recognition. The innovative strategies of CRHM consist of the hierarchical structure, the groups of fuzzy subsystems and the cycle mechanism. The hierarchical structure is innovatively built to extract features and transform the low-level features into advanced ones semantically, in which we adopt the groups of fuzzy subsystems as feature extraction units in each hidden layer, which ensures the diversity of features, avoids the fuzzy rules explosion, and reduces the time for clustering. The cycle mechanism is first proposed to connect the hierarchical structure and the output layer directly, transferring the tuned parameters again and again, to reinforce features gradually. To demonstrate the performance of CRHM, we have conducted extensive comparison with several state-of-the-art algorithms on benchmark 1D and 2D datasets. The experimental results show that the recognition rate of CRHM is higher than convolutional neural network (CNN), while the training time is only 5% of CNN's, which confirms that our approach provides a novel model for recognition, which can simultaneously improve the effectiveness and efficiency without the need of advanced equipment. In addition, the analysis results about the contribution of the core strategies to CRHM performance indicates that the contribution of the hierarchical structure is greater than that of the groups of fuzzy subsystems, which is superior than that of the cycle mechanism. [ABSTRACT FROM AUTHOR]

A black silicon on emitter (BSOE) solar cell, with a uniform depletion region, was investigated by its internal quantum efficiency (IQE) and by scanning electron microscopy using an electron beam-induced current. We establish a three-dimensional energy band diagram of a non-uniform depletion region to illustrate the shunt effect by introducing a lateral electric field into the black silicon solar cell. The IQE spectral response at 450nm is enhanced by ∼31%. The conversion efficiency is improved by 0.8% abs compared with a black silicon solar cell. [Copyright &y& Elsevier]

Abstract: An optimized textured structure is a key for high efficiency multi-crystalline silicon solar cell. In this work, black silicon wafers with various structures have been successfully produced by plasma immersion ion implantation. The surface morphology, reflectance and internal quantum efficiency have been investigated by atomic force microscope, spectrophotometer and quantum efficiency measurement system, respectively. Results show that nanohillocks with average height of 150–600nm have been generated on black silicon surfaces by different texturing conditions, and the reflectance over the wavelength from 300nm to 1100nm decreases with increasing the height of nanohillocks, whereas the internal quantum efficiency worsens. The solar cell based on the optimized nanohillocks height of 300nm yeilds efficiency of 15.99% with short circuit current of 34.0mA/cm2. [Copyright &y& Elsevier]

Abstract: Optimized textured structure is one of the most important elements for high efficiency multicrystalline silicon solar cells. In this paper, in order to incorporate low reflectance nanostructures into conventional industrial solar cells, structures with aspect ratios of about 1:1 and average reflectance of 8.0% have been generated using plasma immersion ion implantation. A sheet resistance of 56.9Ω/sq has been obtained by adjusting the phosphorous diffusion conditions, while the thickness of the silicon nitride vary in 70–80nm by extending the deposition time by 100s. Under the conventional co-firing conditions, a solar cell with efficiency of 16.3% and short-circuit current density 34.23mA/cm2 has been fabricated. [Copyright &y& Elsevier]

Abstract: In the present study, the black silicon has been successfully produced by plasma immersion ion implantation (PIII). The microstructure and the reflectance of the black silicon have been investigated by field emission scanning electron microscope and spectrophotometer, respectively. Results show that the black silicon exhibits a needle-like structure with the average reflectance of 1.79%. The solar cell based on black silicon yields an efficiency of 15.68% with a fill factor (FF) of 0.783. [Copyright &y& Elsevier]

Abstract: It is a major issue to improve the thermo-mechanical sensitivity of uncooled optomechanical focal plane arrays (FPAs) for infrared imaging. This work presents an optimized multi-fold interval metallized leg (IML) configuration to increase the thermo-mechanical sensitivity of an uncooled optomechanical bi-material micro-cantilever array. The inclination angle changes of the cantilever elements are measured in the IR imaging system using an optical readout with a knife-edge filtering operation in the spectrum plane. The multi-fold IML configuration consists of alternately connected unmetallized and metallized legs. With the optimized fold number, the thermo-mechanical sensitivity of a micro-cantilever array can be amplified to two times of one-fold IML for a 120μm×120μm element with 1μm thick SiN x /0.2μm thick Au films. Room temperature objects are imaged with the fabricated FPA containing 160×160 elements and a 12-bit CCD. Further modeling analysis shows that the experimental results are well accordant with the theoretical calculation. An important practical feature of the implemented approach is its straightforward fabrication for a large FPA, without growing complexity and cost. [Copyright &y& Elsevier]

Abstract: In this study, we present an uncooled infrared imaging detector using knife-edge filter optical readout method. The tilt angle change of each cantilever in a focal plane array (FPA) can be simultaneously detected with a resolution of 10−5°. A deformation magnifying substrate-free microcantilever unit is specially designed. The multi-fold legs of microcantilever are interval metal coated to form a thermal deformation magnifying structure. Thermal and thermomechanical performance of this microcantilever unit are modeled and analyzed. An FPA with 100×100 pixels is fabricated and thermal images of human body are obtained by this detector. [Copyright &y& Elsevier]

Abstract: The thermal imaging performance of bi-material microcantilever uncooled infrared focal plane array depends on its thermal deformation sensitivity (the ratio of the microcantilever deformation to a temperature change of thermal objects). In this paper, two microcantilever arrays with different structures are designed and fabricated, which contain multi-fold interval metallized legs. The thermo-inducing incline angle of the microcantilever is measured by knife-edge filtering optical readout method. One of the structures has deformation magnification effect, which can magnify the incline angle of microcantilevers under the same temperature rise. The other structure aims at keeping microcantilevers from the influence of environmental temperature fluctuation. The performance of the microcantilever structures is modeled. Thermal images of human body are obtained with the fabricated microcantilever arrays. The thermal deformation sensitivities of two structures are measured, which agree with the modeling analyses. [Copyright &y& Elsevier]

Abstract: This paper describes a novel single-layer bi-material cantilever microstructure without silicon (Si) substrate for focal plane array (FPA) application in uncooled optomechanical infrared imaging system (UOIIS). The UOIIS, responding to the radiate infrared (IR) source with spectral range from 8 to 14μm, may receive an IR image through visible optical readout method. The temperature distribution of the IR source could be obtained by measuring the thermal–mechanical rotation angle distribution of every pixel in the cantilever array, which is consisted of two materials with mismatching thermal expansion coefficients. In order to obtain a high detection to the IR object, gold (Au) film is coated alternately on silicon nitride (SiN x ) film in the flection beams of the cantilevers. And a thermal–mechanical model for such cantilever microstructure is proposed. The thermal and thermal–mechanical coupling field characteristics of the cantilever array structure are optimized through numerical analysis method and simulated by using the finite element simulation method. The thermal–mechanical rotation angle simulated and thermal–mechanical sensitivity tested in the experiment are 2.459×10−3 and 3.322×10−4 rad/K, respectively, generally in good agreement with what the thermal–mechanical model and numerical analysis forecast, which offers an effective reference for FPA structure parameters design in UOIIS. [Copyright &y& Elsevier]