Presentation Attack Detection Using a Tiny Fully Convolutional Network

Automatic Optical Inspection (AOI) is any method of detecting defects during a Printed Circuit Board (PCB) manufacturing process. Early AOI methods were based on classic image processing algorithms using a reference PCB. The traditional methods require very complex and inflexible preprocessing stages. With recent advances in the field of deep learning, especially Convolutional Neural Networks (CNN), automating various computer vision tasks has been established. Limited research has been carried out in the past on using CNN for AOI. The present systems are inflexible and require a lot of preprocessing steps or a complex illumination system to improve the accuracy. This paper studies the effectiveness of using CNN to detect soldering bridge faults in a PCB assembly. The paper presents a method for designing an optimized CNN architecture to detect soldering faults in a PCBA. The proposed CNN architecture is compared with the state-of-the-art object detection architecture, namely YOLO, with respect to detection accuracy, processing time, and memory requirement. The results of our experiments show that the proposed CNN architecture has a 3.0% better average precision, has 50% less number of parameters and infers in half the time as YOLO. The experimental results prove the effectiveness of using CNN in AOI by using images of a PCB assembly without any reference image, any complex preprocessing stage, or a complex illumination system. Doi: 10.28991/HIJ-2022-03-01-01 Full Text: PDF

Data structures for virtual-circuit implementation

The S-transform provides both the true frequency and globally referenced phase measurements characteristic of the Fourier transform and also generates local spectra, as does the wavelet transform. But the computational requirements of the 2D S-transform have limited its application in digital images processing in time. In order to reduce the processing time and memory requirement, we present a fast oversampling orthogonal method to compute the discrete 2D S-transform. Our method combines Brown's oversampling method with Drabycz's orthogonal method and reduces the time and memory consuming dramatically. We calculated the 2D discrete S-transform of the typical 24 bits bmp image in 0.05 seconds and the memory requirement is 8 times of the original image size. 256×256 24 bits bmp image in 0.05 seconds and the memory requirement is 8 times of the original image size.

The analysis of remote sensing data enables us to detect changes and monitor land surface temperature (LST). However, analysis of times series data poses some challenges, including weather conditions, seasonality and noise, that limit the effectiveness of change detection algorithms. While existing algorithms perform relatively well for detecting abrupt transitions, reliable detection of gradual changes is more difficult. In this paper, we formulate the problem of spatiotemporal LST detection as a statistical sequential change detection problem. LST images are modeled as stochastic processes, with temperature changes reflected as changes in the parameters (i.e., mean) of the process. A generalized likelihood ratio test is used to detect these changes and estimate the exact time/space where they occur. To minimize processing time and memory requirements, we represent LST images by their reduced dimensionality using direct cosine transformation followed by principal component analysis. Statistical sequential analysis is used to provide a unified mathematical framework for the detection of both abrupt and gradual changes in LST observations of Bridgeton Missouri landfill over 17 years.

Linear assignment problems are well-known combinatorial optimization problems involving domains such as logistics, robotics and telecommunications. In general, obtaining an optimal solution to such problems is computationally infeasible even in small settings, so heuristic algorithms are often used to find near-optimal solutions. In order to attain the right assignment permutation, this study investigates a general-purpose learning strategy that uses a bipartite graph to describe the problem structure and a message passing Graph Neural Network (GNN) model to learn the correct mapping. Comparing the proposed structure with two existing DNN solutions, simulation results show that the proposed approach significantly improves classification accuracy, proving to be very efficient in terms of processing time and memory requirements, due to its inherent parameter sharing capability. Among the many practical uses that require solving allocation problems in everyday scenarios, we decided to apply the proposed approach to address the scheduling of electric smart meters access within an electricity distribution smart grid infrastructure, since near-real-time energy monitoring is a key element of the green transition that has become increasingly important in recent times. The results obtained show that the proposed graph-based solver, although sub-optimal, exhibits the highest scalability, compared with other state-of-the-art heuristic approaches. To foster the reproducibility of the results, we made the code available at https://github.com/aircarlo/GNN_LSAP.

An approximation heuristic is proposed for solving the heterogeneous multi-server queueing problem associated with the analysis of resequencing of packets travelling over multiple physical links in a packet-switching network. Even though a method for obtaining the exact solution exists, its processing time and memory requirements vary exponentially in terms of the number of servers and render it infeasible even for moderately-sized systems. Precision of the proposed approximation which has linear time complexity is demonstrated. The approximation is recommendable in cases when system population and resequencing delays, rather than individual link utilizations, have to be calculated. >

Carpooling is an effective solution to traffic congestion. It increases the usage rate of vehicles by employing empty seats as a transportation resource. In order to provide carpooling services to users, we developed an intelligent carpool system called BlueNet-Ride . After prospective carpoolers submit their requests through their smart handheld devices, this system provides appropriate matches by using the proposed Low-Complexity and Low-Memory Carpool Matching method. The compact genetic algorithm is applied to our Low-Complexity and Low-Memory Carpool Matching method, which involves three proposed modules: an Evolutionary Model Initialization module, an Evolutionary Process Operation module, and an Evolutionary Model Modification module. The Evolutionary Model Initialization module takes advantage of the manipulation of the evolving population on a probability distribution to achieve low-memory requirements during the evolution process of the carpool match solution. The Evolutionary Process Operation and Evolutionary Model Modification modules simulate genetic operations to accomplish superior matching within a short amount of time. The experimental results demonstrate that our Low-Complexity and Low-Memory Carpool Matching method achieves the highest degree of performance with regard to solution quality, processing time, and memory requirements of all evaluated methods.

<span lang="EN-US">The revolution in 5G mobile systems require changes to how image is handled. These changes are represented by the required processing time, the amount of space for uploading and downloading. In this paper, a development on WT (Wavelet Transform) along with LM-SPIHT (Listless-Modified Set Partitioning in Hierarchical tree) coding and with additional level of Runlength encoding for image compression has been proposed. The new implementation reduces the amount of data needed to be stored in several stages, also the amount of time required for processing. The compression has been implemented using VHDL (Very High Descriptive Language) on netFPGA-1G-CLM Kintex-7 board. The new implementation results show a reduction in the complexity as processing time.</span>

Fingerprint recognition on smartphone provides a good alternative over traditional security measures such as lock-patterns and pin. However, such fingerprint systems have some inherent problems such as the fact that user will leave their latent fingerprint on the sensor, and the limited sensor area. Additionally, fingerprint sensors impact on the cost and form factor of the device. Hence, in literature, the camera based approaches such as fingerphoto recognition systems got the attention of many researchers and manufacturers. However, such systems are highly vulnerable to presentation attacks such as photo-prints, display and replay attacks. To countermeasure these attacks, we propose a robust presentation attack detection scheme based on the features extracted from the maximum response images obtained from the convolution of second order Gaussian derivatives and the input images at multiple scales. The proposed scheme has achieved the detection performance of BPCER of 1.8%, 0.0% and 0.66% at APCER=10% for the presentation attack instrument species i.e., print-photo, display and replay attacks respectively.

It has been shown that time series containing reflectance values from the first two spectral bands of the MODerate-resolution Imaging Spectroradiometer (MODIS) land surface reflectance product can be modelled as a triply modulated cosine function. A meta-optimization approach has been proposed in the literature for setting the parameters of the non-linear Extended Kalman Filter (EKF) to rapidly and efficiently estimate the features for these triply modulated cosine functions using spatial information. In this paper we modify this approach to utilize temporal information instead of spatial information to greatly reduce the processing time and storage requirements to process each time series. The parameters derived from the newly proposed method is classified with a support vector machine and compared to the original approach. Performance of the methods is tested on the Limpopo province in South Africa.

Reconstructing damaged images is a critical challenge in image processing, with applications ranging from historical restoration to medical imaging. This paper presents a novel approach that combines the Pix2Pix network with an attention mechanism to improve the reconstruction of damaged images. Initially, images were resized and categorized into grayscale and vintage styles, with artificial damages applied. The Pix2Pix network was employed both independently and in conjunction with the attention mechanism for reconstruction. Our results indicate that integrating the attention mechanism significantly enhances reconstruction quality, particularly in terms of detail and clarity. Evaluations based on PSNR and SSIM metrics demonstrate the superiority of the combined model. This method holds promise for various applications, including the restoration of historical artifacts and medical imaging, although challenges such as processing time and data requirements remain. Future work will focus on refining the model and exploring data augmentation techniques to further improve performance.

In many data stream mining applications, traditional density estimation methods such as kernel density estimation, reduced set density estimation can not be applied to the density estimation of data streams because of their high computational burden, processing time and intensive memory allocation requirement. In order to reduce the time and space complexity, a novel density estimation method Dm-KDE over data streams based on the proposed algorithm m-KDE which can be used to design a KDE estimator with the fixed number of kernel components for a dataset is proposed. In this method, Dm-KDE sequence entries are created by algorithm m-KDE instead of all kernels obtained from other density estimation methods. In order to further reduce the storage space, Dm-KDE sequence entries can be merged by calculating their KL divergences. Finally, the probability density functions over arbitrary time or entire time can be estimated through the obtained estimation model. In contrast to the state-of-the-art algorithm SOMKE, the distinctive advantage of the proposed algorithm Dm-KDE exists in that it can achieve the same accuracy with much less fixed number of kernel components such that it is suitable for the scenarios where higher on-line computation about the kernel density estimation over data streams is required.We compare Dm-KDE with SOMKE and M-kernel in terms of density estimation accuracy and running time for various stationary datasets. We also apply Dm-KDE to evolving data streams. Experimental results illustrate the effectiveness of the proposed method.

Fast monochromatic radiative transfer calculations for limb sounding

Energy harvesting sensors move into the mainstream because they are regarded as a perfect match for the monitoring applications. Nonetheless, their implementation is only feasible if the computing unit, generally a microcontroller, is capable of running the real-time tasks using sophisticated scheduling and power management techniques. In this paper, we address the hard real-time scheduling problem of Directed Acyclic Graph (DAG) tasks in a uniprocessor RTEH (Real Time Energy Harvesting) computing system. In this model, a task is defined as a set of dependent sub-tasks that execute under precedence constraints with processing time and energy requirements. We demonstrate that the energy aware ED-H algorithm provides an optimal solution. We show that the scheduling policy directly includes a feasibility analysis by enforcing the precedence constraints, thus making it possible to get rid of semaphore synchronization.

Enablers to computer vision technology for sustainable E-waste management