Rectangular Features Research Articles

Facial Frontiers: Unveiling the Potential of LBPHs and Haar Cascades in Facial Recognition for Enhanced School Security

The article aims at introducing a Facial Recognition School Security System as one solution to improve security while speeding up administrative processes in educational institutions. The need for such a system emanates from concerns over the safety of schools as well as inadequacies inherent in conventional attendance and access control methods. Inadequate methods can be manual or rely on an older technology that leads to inefficiencies, inaccuracies and breaches of security. This proposed solution exploits contemporary artificial intelligence algorithms and computer vision techniques to facilitate the reliable identification and validation of entrants thereby providing contactless approach during Pandemic times in compliance with the COVID-19 safety protocols. At this period of COVID-19 it also acts as an avenue where physical touchpoints are reduced with consideration to social distancing measures. Our method is novel in the fact that it only detects and recognizes human faces as opposed to general object detection systems. We use Local Binary Pattern Histograms (LBPH) for face recognition and Haar Cascades for face detection. The Haar Cascade algorithm employs simple rectangular features to detect faces, using a cascade of weak classifiers to achieve high detection rates. The LBPH algorithm captures local texture patterns of facial features, calculating LBP values for each pixel. Our project demonstrates variable performance across different classes, with precision ranging from 0.50 to 1.00, recall from 0.33 to 1.00, and F1 scores from 0.33 to 0.94, while achieving an overall accuracy of 0.75, indicating robust performance in certain scenarios but room for improvement in others.

SMART materials for DEMO: Towards industrial production

Self-passivating Metal Alloys with Reduced Thermo-oxidation (SMART) are under development for a fusion power plant. SMART exhibits similar sputtering resistance as that of pure tungsten during regular plasma operation. Under accident conditions, SMART demonstrates the suppression of oxidation – a remarkable advantage over pure tungsten. The viability of SMART technology has been shown on a laboratory scale.Presently, the industrial scale-up of SMART technology is underway. This activity comprises the industrially supplied feedstock materials for SMART, mechanical alloying of powder at industry and sintering of SMART materials using industrial equipment. Industrial mechanical alloying can be accomplished within 21,5 h of milling, providing at least four kilograms of fully alloyed powder. Rectangular industrial SMART samples feature linear dimensions of 10 cm, a thickness of 0.5 cm, a weight slightly below 1 kg and a relative density exceeding 97 %.The transition to industrial feedstock suppliers allowed SMART to outperform pure tungsten in powder procurement costs needed for plasma-facing components. The status of industrial SMART technology is given along with an outlook on future activities.

Optimal resource placement for electric grid resilience via network topology

In this paper, we investigate the resilience of alternative electric grid configurations by adopting a stylized approach based on graph theory, probabilistic analysis, and simulation. We consider two alternative classes of electricity network topology: binary trees and rectangular lattices. For each topology, we find the probabilities that customers located at various nodes in the network will continue to have power following a disaster, depending on the locations of resources (e.g., generators, storage units) in the network. Then, these probabilities are incorporated into the problem of optimally placing resources throughout the network. This is a cost–benefit problem that weighs the benefits of placing resources closer to customers – that is, pursuing a distributed resilience strategy – against the higher total cost of deploying a greater number of smaller resource units. Our analytical and numerical results thus shed light on the general circumstances in which centralized or distributed resilience strategies are preferable. While optimal resource placements depend on various assumptions, such as the probability that power lines fail and the strength of economies of scale, we find that distributed resilience strategies are more often preferred in the binary tree topology than in the rectangular lattice topology. Rectangular lattices feature greater redundancy in terms of paths between nodes in the network, enabling the system to be fairly resilient even with centralized resources.

Rectangular Natural Feature Recognition and Pose Measurement Method for Non-Cooperative Spacecraft

Accurately estimating the pose of spacecraft is indispensable for space applications. However, such targets are generally non-cooperative, i.e., no markers are mounted on them, and they include no parts for operation. Therefore, the detection and measurement of a non-cooperative target is very challenging. Stereovision sensors are important solutions in the near field. In this paper, a rectangular natural feature recognition and pose measurement method for non-cooperative spacecraft is proposed. Solar panels of spacecraft were selected as detection objects, and their image features were captured via stereo vision. These rectangle features were then reconstructed in 3D Cartesian space through parallelogram fitting on the image planes of two cameras. The vertexes of rectangle features were detected and used to solve the pose of a non-cooperative target. An experimental system was built to validate the effectiveness of the algorithm. The experimental results show that the average position measurement error of the algorithm is about 10 mm and the average attitude measurement error is less than 1°. The results also show that the proposed method achieves high accuracy and efficiency.

Advancements in CNC Machining: Rectangular Boss Feature Validation and Toolpath Precision using STEP AP242 and B-Rep Data Structure

In the context of a CNC machining application, this study proposes an alternate approach for creating toolpaths designed explicitly for machining rectangular bosses utilizing the STEP AP242 file format as a data source. The suggested methodology ensures data accuracy and integrity maintenance by extracting exact geometric information from digital models using Geometric Data Extraction (GDE) tools. The value of the Cartesian multi-point decimal point in the data structure and variations in the reference number of the Cartesian point every time a user accesses and converts to a STEP (AP203) file are two drawbacks of AP203 when compared to the widely utilized AP203 in prior technique studies. This method makes it possible to create G-code efficiently, which is necessary for managing computer numerical control (CNC) machines, by utilizing the power of STEP AP242 and GDE. The process entails finding and separating the rectangular island milling features in the STEP AP242 files, taking out critical geometric characteristics, and creating toolpaths that are optimal for CNC machining. Through this approach, the final G-code instructions are guaranteed to be precisely calibrated to consider every island feature's distinct quality while preserving overall machining efficiency. Several real-world case studies and valuable applications are investigated to confirm the efficacy of the suggested method. The results show significant increases in productivity, decreased waste material, and improved precision of CNC machining, all of which contribute to a more effective and economical production process.

Szolnok – Milléri-holtág part I-II. több korszakú lelőhely őrlőkő leletei archeometriai vizsgálatának előzetes eredményei : Új adatok a domoszlói piroxénandezit nyersanyag régészeti elterjedéséhez

The multi-period archaeological site Szolnok – Milléri-holtág part I-II is located in the north-eastern vicinity of Szolnok, approximately 5 km from the town, on the high bank of the Milléri-holtág. The total excavated area is almost 27 000 m2 . The preventive excavation was performed by the staff of the Hungarian National Museum in two campaigns (in 2014 and 2015) connected with the M4 motorway constructions. Most excavated settlement features belonged to a Middle Iron Age (Scythian Period) and a Late Iron Age (Celtic Period) rural settlement with scattered structures. Besides these, other settlement features were found from the Neolithic, Late Bronze Age and Medieval Period (Árpád Period). It has to be mentioned that in the excavated part of the Scythian Period settlement, only circular and oval structures were found. The rectangular features – typical for other settlements of the period – were absent in Milléri-Holtág part I-II. However, the archaeological finds were typical: ceramic assemblages were characterised by long-lasting forms (pots, bowls) of the Iron Age household ceramics. The assemblages of the Late Iron Age features were very similar to the vessel types of the earlier period. The partition of the artefacts can only be made with the help of some characteristic forms and ornamentations that appeared in a close unit. The ground stone assemblage of the site (primarily fragments of grinding stones) was analysed with petrographic and geochemical methods. Based on these investigations, the finds' most frequent raw material type could be identified with the Domoszló-type pyroxene andesite from the Mátra Mountains, studied intensively in the past few years. Other types of andesite were present in the ground stone assemblage in smaller amounts. The identification of the source areas of these different andesite raw material types needs more exploration, not just in the territory of Hungary, but also in some areas beyond the borders, e.g. Subcarpathia (Zakarpatska Oblast, Ukraine), Eastern-Slovakia, Transylvania (Romania).

Array configuration-agnostic personalized speech enhancement using long-short-term spatial coherence.

Personalized speech enhancement (PSE) has been a field of active research for suppression of speech-like interferers, such as competing speakers or television (TV) dialogue. Compared with single-channel approaches, multichannel PSE systems can be more effective in adverse acoustic conditions by leveraging the spatial information in microphone signals. However, the implementation of multichannel PSEs to accommodate a wide range of array topology in household applications can be challenging. To develop an array configuration-agnostic PSE system, we define a spatial feature termed the long-short-term spatial coherence (LSTSC) with a dynamic forgetting factor as the input feature to a convolutional recurrent network to monitor the spatial activity of the target speaker. As another refinement, an equivalent rectangular bandwidth-scaled LSTSC feature can be used to reduce the computational cost. Experiments were conducted to compare the proposed PSE systems, including the complete and the simplified versions with four baselines using unseen room responses and array configurations (geometry and channel count) in the presence of TV noise and competing speakers. The results demonstrated that the proposed multichannel PSE network trained with the LSTSC feature with a dynamic forgetting factor achieves superior enhancement performance without precise knowledge of the array configurations and room responses.

Mechanisms of Plasma Actuators Controlling High-Aspect-Ratio Rectangular Jet Width for Automobile Air Conditioning Systems

High-aspect-ratio (HAR) rectangular jets have attracted attention in automobile air conditioning (A/C) systems and turbulent jet applications owing to their excellent air delivery and mixing and attractive interior design. Active flow control (AFC) of rectangular jets using plasma actuators (PAs) has proven to be a promising technique because the actuator is simple, has low energy consumption, and can create flow features without interference. This research aims to understand the interaction between PAs and flow from a HAR rectangular nozzle using hot-wire anemometry, particle image velocimetry, and theoretical studies. Understanding how PAs affect the flow is beneficial for designing air vents to fit automobile A/C systems and various engineering applications by recreating the flow features with other AFC techniques and actuators. The combination of periodic excitation and vectoring effects transfers the flow’s mean energy to organized structures—known as spanwise vortexes—as large as 6 mm. The interaction between these coherent structures and the dissipative environment compresses the vortexes, resulting in the flow converging on the spanwise–streamwise (X–Z) plane and diverging on the transverse–streamwise (X–Y) plane. HAR rectangular jet flow features controlled by PAs can be predicted for specific cases by calculating the Strouhal number based on PA operating parameters.

Observation of gapped Dirac cones in a two-dimensional Su-Schrieffer-Heeger lattice

The Su-Schrieffer-Heeger (SSH) model in a two-dimensional rectangular lattice features gapless or gapped Dirac cones with topological edge states along specific peripheries. While such a simple model has been recently realized in photonic/acoustic lattices and electric circuits, its material realization in condensed matter systems is still lacking. Here, we study the atomic and electronic structure of a rectangular Si lattice on Ag(001) by angle-resolved photoemission spectroscopy and theoretical calculations. We demonstrate that the Si lattice hosts gapped Dirac cones at the Brillouin zone corners. Our tight-binding analysis reveals that the Dirac bands can be described by a 2D SSH model with anisotropic polarizations. The gap of the Dirac cone is driven by alternative hopping amplitudes in one direction and staggered potential energies in the other one and hosts topological edge states. Our results establish an ideal platform to explore the rich physical properties of the 2D SSH model.

Novel quantitative electroencephalogram feature image adapted for deep learning: Verification through classification of Alzheimer's disease dementia.

Quantitative electroencephalography (QEEG) analysis is commonly adopted for the investigation of various neurological disorders, revealing electroencephalogram (EEG) features associated with specific dysfunctions. Conventionally, topographies are widely utilized for spatial representation of EEG characteristics at specific frequencies or frequency bands. However, multiple topographies at various frequency bands are required for a complete description of brain activity. In consequence, use of topographies for the training of deep learning algorithms is often challenging. The present study describes the development and application of a novel QEEG feature image that integrates all required spatial and spectral information within a single image, overcoming conventional obstacles. EEG powers recorded at 19 channels defined by the international 10-20 system were pre-processed using the EEG auto-analysis system iSyncBrain®, removing the artifact components selected through independent component analysis (ICA) and rejecting bad epochs. Hereafter, spectral powers computed through fast Fourier transform (FFT) were standardized into Z-scores through iMediSync, Inc.'s age- and sex-specific normative database. The standardized spectral powers for each channel were subsequently rearranged and concatenated into a rectangular feature matrix, in accordance with their spatial location on the scalp surface. Application of various feature engineering techniques on the established feature matrix yielded multiple types of feature images. Such feature images were utilized in the deep learning classification of Alzheimer's disease dementia (ADD) and non-Alzheimer's disease dementia (NADD) data, in order to validate the use of our novel feature images. The resulting classification accuracy was 97.4%. The Classification criteria were further inferred through an explainable artificial intelligence (XAI) algorithm, which complied with the conventionally known EEG characteristics of AD. Such outstanding classification performance bolsters the potential of our novel QEEG feature images in broadening QEEG utility.

Implementasi Algoritma Haar Cascade Classifier Dalam Mendeteksi Robot Sepak Bola Beroda

In soccer robot contests, generally soccer robots can recognize their own team robots through color detection using the HSV model. Robots that use color detection to identify their own team robot can detect objects that have the same color value, so objects that have the same color will be considered as their own team robot. However, if you only rely on this method, it is still lacking when viewed in terms of object tracking. Haar like feature or known as Haar Cascade Classifier is a rectangular (square) feature method, which gives a specific indication of an image. This method is able to detect quickly and in real time. Therefore, the researchers tried to unite these two detections to be able to produce a better detection system than before. It is hoped that hoped that in this study, researchers can provide input for the Robotics team of Dinamika Bangsa University in improving the accuracy of robot detection against robots in the same team. From the results of the study it was found that the detection of objects with a distance from the camera as far as 100 cm to 150 cm was successfully carried out, but with a distance of more than 150 cm the object was not detected.

Unsupervised feature extraction with autoencoders for EEG based multiclass motor imagery BCI

Decoding of motor imagery (MI) from Electroencephalogram (EEG) is an important component of BCI system that helps motor-disabled people interact with the outside world via external devices. One of the main issues associated with the multiclass classification of MI based EEG is the informative confusion due to non-stationary characteristics of EEG data. In this work, an innovative idea of transforming EEG signal into a new domain, weight vector of autoencoder, unsupervised neural network, is proposed for the first time to solve that confusion. These weight vectors are optimized according to that particular EEG signal. The features: autoregressive coefficients (ARs), Shannon entropy (SE) and wavelet leader were extracted from the weight vector. A rectangular windowing-based feature extraction technique is implemented to capture the local features of the EEG data. Finally, extracted features were used in the support vector machine (SVM) as a classifier network. The proposed method is implemented on two openly available EEG dataset (BCI competition-III and Competition-IV) to validate the effectiveness and superiority of the proposed methodology over the newly reported methods. For four-class EEG based MI classification, the proposed technique has achieved an average test accuracy of 95.33% and 97% for dataset-IIIa from BCI-III and dataset-IIa from BCI-IV respectively. The experimental results reveal that, the proposed technique is a promising solution to improve the decoding performance of BCIs.

Colourful treasure on Shanghai’s Bund: uncovering original paint finishes in the Chinese Banking Hall of the former HSBC Building

The former HSBC Building, today occupied by Shanghai Pudong Development Bank, is one of the largest iconic buildings on the Bund in Shanghai. Designed by Palmer & Turner in the neoclassic style, the building was completed in 1923. Restoration of the interior of the Chinese Banking Hall, located within the southwestern corner of the building, is now underway. This rectangular room features four columns in the centre and stucco decoration on parts of the walls; the ceiling originally had a distinctive, vibrantly colourful design in a bold mix of Western and Chinese styles. The only extant documentation of the banking hall’s original decoration includes black-and-white photographs and a short, written description. Today, all the plaster surfaces are covered over with several layers of white and cream-coloured paint finishes. A preliminary analysis of the hall’s decorative finishes was conducted to determine the colours, paint materials, and condition of the original paint layers. The investigation revealed that in some sections, the original paint was very brittle. Restoration by exposing the finishes in the entire room (removing the modern layers covering the original one) did not appear feasible. This article investigates the results of the following: assessing the condition of the paint materials; identifying the significance of the decorations; the challenges facing the room’s conservation; and potential conservation options for the room’s decoration.

OSO-YOLOv5: Automatic Extraction Method of Store Signboards in Street View Images Based on Multi-Dimensional Analysis

To realize the construction of smart cities, the fine management of various street objects is very important. In dealing with the form of objects, it is considered a pursuit of normativeness and precision. Store signboards are a tangible manifestation of urban culture. However, due to factors such as high spatial heterogeneity, interference from other ground objects, and occlusion, it is difficult to obtain accurate information from store signboards. In this article, in response to this problem, we propose the OSO-YOLOv5 network. Based on the YOLOv5 network, we improve the C3 module in the backbone, and propose an improved spatial pyramid pooling model. Finally, the channel and spatial attention modules are added to the neck structure. Under the constraint of rectangular features, this method integrates location attention and topology reconstruction, realizes automatic extraction of information from store signboards, improves computational efficiency, and effectively suppresses the effect of occlusion. Experiments were carried out on two self-labeled datasets. The quantitative analysis shows that the proposed model can achieve a high level of accuracy in the detection of store signboards. Compared with other mainstream object detection methods, the average precision (AP) is improved by 5.0–37.7%. More importantly, the related procedures have certain application potential in the field of smart city construction.

Evaluation of Electrochemical Performance Tuning By Laser Structuring of Electrodes and Its Impact on Cell Degradation Mechanisms

Tuning of electrochemical properties of lithium-ion batteries by using ultrafast laser structuring of electrodes is a rather new technical approach. At Karlsruhe Institute of Technology (KIT) this technology was developed and established for the first time. However, quite recently, other research groups and institutions worldwide also recognized the high potential of this technology for lithium-ion battery production. Lately, a roll-to-roll laser processing infrastructure for structuring large footprint electrode materials for high capacity pouch cells and cylindrical cells was installed at KIT in order to push this technology to demonstrator level, namely technology readiness level 6. The high rate capability of the produced high energy battery cells were significantly improved and twice lifetime of cells could be achieved in comparison to cells with unstructured electrodes. In frame of research, cooperative, and industrial projects the impact of laser structuring for small and large footprint electrodes was studied with regard to lithium distribution caused electrochemical cycling. Hereby, different types of LiNixMnyCo1-x-yO2 (NMC) as a cathode material and graphite as well as silicon-graphite as an anode material with areal capacities up to 4 mAh/cm2 were investigated. Quantitative lithium distribution along entire electrodes were measured by laser-induced breakdown spectroscopy (LIBS). From the 3D elemental mapping starting points for electrochemical degradation could be identified. In cathode materials the local increase of lithium indicated the formation of electrical short cuts which were most related to electrode material inhomogeneity, e.g., by a macroscopic change in porosity or macroscopic film defects. The compressive stress applied to electrodes and separator during cycling has also a significant influence on lithium distribution and subsequent cell degradation. The appropriate type of laser generated patterns such as holes, lines or grids strongly depends on the application scenario as well as type of materials. For silicon-graphite anodes the huge volume expansion during lithium silicide needs to be considered and more rectangular and broader structure features become relevant. Laser structuring of electrodes showed in all cases, for small and large footprint cathodes and anodes, a rather homogenized lithium distribution along the surface and electrode thickness. For full cells elemental mapping of electrodes facing each other was performed indicating the mutual influence on lithium distribution. It can be concluded that the laser-based 3D electrode concept is beneficial regarding a reduced cell degradation. In addition, it could be proven by LIBS that for the 3D battery concept new lithium diffusion pathways were generated which becomes activated at elevated C-rates. Those new lithium diffusion pathways counteract the increase of diffusion overpotential leading to a higher capacities also for high power operation.

New Perspective on Neolithic Rectangular Features Using Artefact Analysis, Soil Micromorphology and Ethnohistorical Analogies. A Case study from Střelice u Brna, South Moravia, Czech Republic

Abstract The function of sunken rectangular features in the Late Neolithic has been a subject of interest in Central Europe for many years. This type of pit is not found very often. One of the latest examples is a square Lengyel feature in the village of Střelice near Brno. A study of the microstratigraphy of the fill has made it possible to qualitatively move the debate to a new level. The qualitative assessment of bone micro-fragments is also innovative for the method of micromorphology in the archaeological context. The study also includes an analysis of artefacts in the actual fill of feature 562, especially the deposits of fragments of “other lithic industry”, pottery and parts of an animal’s body. The discovered situation is then interpreted in the context of other known contemporary features of this type. The presence of excrement with digested bone micro-fragments in the lower layer and the absence of a floor treatment, tread horizon or bedding can be interpreted in the context of ethnohistorical analogies, e. g., as the short-term use of a faecal pit for an omnivorous or carnivorous species (bear) with a subsequent ritual ending.

Exploring the nature of buried linear features in the Qatar peninsula: Archaeological and paleoclimatic implications

The absence of large archaeological settlements in the Eastern Arabian Peninsula suggests that most of the recent wet periods occurring during climate oscillations in these areas were short-lived. However, a ∼ 2-by-3-km buried rectangular feature was discovered opportunistically in the southeast of the Qatar Peninsula while performing radar subsurface mapping of structural conduits for groundwater characterization. The feature is characterized by its unique contrasting patterns of surface and subsurface roughness as well as its organic-rich mixture of soils, as assessed from high-resolution radar and multispectral observations, respectively, and validated with GPR and EM subsurface probing. The above characteristics could be the only remaining evidence for a large primitive settlement due to the harsh erosive setting. To explore the natural or anthropogenic origin of this anomalous feature, we compare its dimensions, structural setting, roughness, and composition to similar formations in the Qatar Peninsula and other arid karst environments. Findings suggest that the feature is texturally and compositionally unique within the Qatar Peninsula, and is not reproduced in other arid karst terrains nor reproduced by the fracture patterns associated with the Qatar-South Fars Arch. We hypothesize that the feature is a karst depression modified by semi-stationary primitive human settlements ∼ 3640 BP or earlier based on radiocarbon dating of in-situ charcoal samples. Moreover, the organic-rich mixture of soil and the occurrence of several symmetrical features observed inside the perimeter suggest that landscaping, farming and irrigation activities could have occurred within the site, further suggesting that long-lasting wet conditions—with sufficient fresh water resources or reliance on karst aquifers—were able to sustain a relatively large primitive settlement not previously witnessed in the region’s archaeological record. Future improvements in the resolution and coverage of orbital L-band radar observations, as used above, can provide new insights into understanding prehistoric human occupation under fluctuating climatic conditions and groundwater dynamics in these hyper-arid regions.

On the Determination of the Chip Nozzle Recognition System by Using Machine Vision

To solve the problem of chip damage caused by the using the wrong type of vacuum nozzle during the packaging of semiconductor chips. A recognition system of vacuum nozzle based on machine vision was proposed. In this research, 29 kinds of lifting nozzles are selected as test samples. The backlight intensity of two lifting nozzle images (one strong and one weak separately) is collected at the first beginning. Then, the Blob analysis method is using to analyze the weak backlighting image. The area of the lifting nozzle and the minimum outer rectangular feature can be obtained subsequently. To identify the shape of the liftin nozzle (round or square), the area ratio is calculated. At the same time, the minimum outer rectangular of the lifting nozzle is selected as the reference rectangle. Then, construct the measurement rectangle. The 2-dimensional size of the lifting nozzle is measured as well. Meanwhile, for the strong backlight image, the average value of the grayscale which located within the minimum outer rectangle is calculated. Therefore, the color (black, white, or beige) of the nozzle can be identified. Finally, the sample data is saved to the database as the sample database. During the recognition process, the shape, color, and size of the lifting nozzle being analyzing are using as the parameter to realize the condition inquire. The experimental results show that the recognition accuracy of this method is 98.85%, and the recognition time of one nozzle is around 1 second, which meets the requirements of practical application.

Buckling resistance of an X80 steel pipeline at corrosion defect under bending moment

This work investigated the resistance of a corroded X80 steel pipe to local buckling at a corrosion defect under bending moment. A nonlinear finite element model was developed to study buckling behavior of the pipe and determine critical bending moment. Parametric effects, including geometry of the corrosion defect (shape, depth, length and width), pipe geometry (pipe outer diameter and wall thickness) and operating pressure, were determined. Results show that pipeline buckling at the corrosion defect depends on the defect geometry. A rectangular corrosion feature results in a lower buckling resistance than a semispherical corrosion pit or a cylindrical grooved corrosion. The depth and width of the corrosion defect are primarily factors affecting the buckling resistance, while the effect of defect length is negligible. The critical bending moment to cause buckling of the pipe, i.e., the critical buckling moment, decreases as the defect depth and width increase. The buckling resistance of the corroded pipe can be improved by increasing the pipe outer diameter and wall thickness. The role of internal pressure in buckling resistance depends on whether a constraint condition is applied at both ends of the pipe. For buried pipelines without sealing ends, the critical buckling moment decreases as the internal pressure increases.

Facial expression recognition using frequency multiplication network with uniform rectangular features

Facial expression recognition (FER) is a popular research field in cognitive interaction systems and artificial intelligence. Many deep learning methods achieve outstanding performances at the expense of enormous computation workload. Limiting their application in small devices or offline scenarios. To cope with this drawback, this paper proposes the Frequency Multiplication Network (FMN), a deep learning method operating in the frequency domain that significantly reduces network capacity and computation workload. By taking advantage of the frequency domain conversion, this novel deep learning method utilizes multiplication layers for effective feature extraction. In conjunction with the Uniform Rectangular Features (URF), our method further improves the performance and reduces the training effort. On three publicly available datasets (CK+, Oulu, and MMI), our method achieves substantial improvements in comparison to popular approaches.