Journal of Progress in Engineering and Physical Science

Creating AI Assistants with Spectral Consciousness Based on Electromagnetic Waves

Evgeny Bryndin — 2026-07-07

The creation of AI assistants with spectral consciousness based on electromagnetic waves relies on periodic functions that can be expanded in a Fourier series and represented by a Fourier integral. Harmonic analysis of the periodic functions is performed to represent them as the sum of a mean value and a set of sinusoidal terms with a fundamental frequency of 1/T and harmonic frequencies of 2/T...n/T electromagnetic waves. Morphogenic cryptocoding of the sinusoidal harmonics is performed based on mental mathematics. Then, cognitive morphogenic modeling of reasoning using harmonic electromagnetic waves is performed according to a standard linguistic template of the judgment process. Based on this modeling, an AI learning model with a Fourier support function is created to create intelligent assistants with spectral consciousness for the development of new knowledge through cognitive verbal and written communication with a specialist and for training in various fields of knowledge and life.

An Improved ShuffleNet with SE Attention and Focal Loss for Lightweight Facial Expression Recognition

Yuqi Zeng — 2026-07-07

Facial expression recognition (FER) plays a critical role in human-computer interaction, mental health monitoring, intelligent security inspection and classroom emotional computing. Traditional deep convolutional neural networks for FER suffer from excessive computational overhead and large parameter volume, which limits their deployment on mobile terminals and embedded devices with limited computing resources. As a classic lightweight network, ShuffleNet adopts group convolution and channel shuffle operations to reduce model complexity, yet it lacks the ability to capture subtle facial emotional features and cannot well handle the class imbalance problem existing in mainstream FER datasets. To address these defects, this paper proposes an improved lightweight ShuffleNet model for facial expression detection. First, Squeeze-and-Excitation (SE) channel attention modules are embedded after each shuffle block to dynamically recalibrate feature channel weights and strengthen the extraction of discriminative facial expression features. Second, a multi-dimensional data augmentation strategy combining geometric transformation, color jitter and Gaussian noise injection is designed to expand sample diversity and enhance the model’s generalization ability under complex lighting, shooting angles and background interference. Third, Focal Loss is introduced to replace the standard cross-entropy loss, which suppresses the loss contribution of easy-classified majority samples and forces the model to focus on hard-to-distinguish minority facial expressions such as anger and disgust. Comprehensive experiments are conducted on three public datasets FER2013, CK+ and AffectNet. The results demonstrate that the proposed model achieves higher recognition accuracy compared with original ShuffleNet V2 while maintaining low computational cost and small parameter size, and presents superior robustness against unbalanced data and complex real-world scenes.

The Four Necessary and Sufficient Canonical Tests of the Special Theory of Relativity

Adrian Sfarti — 2026-07-07

The present paper is a continuation of our previous published work where we dispelled the misconception that the three canonical experiments mentioned by H.P. Robertson in his 1949 seminal paper, are all that is necessary and sufficient in deciding if and when an alternative theory of relativity is equivalent to Einstein’s special relativity.

Anthropometric Studies of the Diurnal Human Height Shrinkage Across Selected Professionals in Nigeria

Mojisola A. Bolarinwa — 2026-07-14

Diurnal variation in stature is one potentially significant source of error in the evaluation of short-term growth. Hence, there is the need to observe whether there are diurnal variations when assessing humans’ (subjects) body composition, using physical anthropometric methods. In this paper, the diurnal human height shrinkage have been investigated by measurements, such that twice-daily measurements and body weight of subjects were examined and recorded. A total of 70 subjects were examined altogether. This comprises of 35 males and 35 females of which 2 nursery school children (a boy and a girl) were involved, aging 4 years and 3½ years respectively. For most of the subjects, Morning measurements were made immediately after rising. The time of the measurements was also recorded and diurnal changes during daytime were estimated. Results showed that stature decreased during the day in the entire 70 subjects surveyed and sampled. From the results, it could be seen that an average female subject will shrink by 0.88% of the usual (morning) height. Also, the mean average shrinkage in height amounts to 1.43 ± 2cm for the females. Likewise, an average male subject will shrink by 0.91% of his usual morning height. The mean average shrinkage in height in male subjects equals 1.56 ± 2cm. The findings of this study confirm the existence of diurnal variation and that the greater proportion of the height loss occurs during the earlier part of the day, especially when subjects are engaged with some form of activity or the other.

Thermo-Mechano-Electromagnetic Multi-Physics Coupling Analysis, Unified-Datum Tolerance Coordination and Full-Lifecycle Experimental Verification of Triple-Functional Integrated Automotive Laminated Windshield Glass

Yan Jiang — 2026-08-04

Intelligent new-energy vehicles increasingly require front windshield laminated glass to integrate electrochromic dimming, low-temperature rapid defrosting and an embedded digital-television (DTV) antenna. The superposition of multilayer polymer films, screen-printed silver conductive traces and embedded flat wiring harnesses, however, introduces thermo-mechano-electromagnetic coupling failure modes. Here, we propose three hierarchically optimized substructures—a stepped-gradient ceramic anti-corrosion shielding layer, a 0.2 mm-protrusion full-wrap harness bonding configuration and full-wrap silicone sealing for marginal un-melted polyvinyl butyral (PVB) zones—and couple them with an ISO 5459-conformant unified symmetric datum (PD-A) tolerance chain model. A fully coupled thermo-mechano-electromagnetic finite-element numerical model, validated through a four-level mesh-convergence study and a five-factor parametric sensitivity matrix, is developed to quantify interfacial stress evolution and electromagnetic interference attenuation. A coupon-grade multi-station test fixture (denoted 8S10Z-39J-900) is fabricated and used for systematic cleavage peeling, alternating thermal cycling, hygrothermal/UV accelerated aging, optical modulation, electrothermal defrost and antenna electromagnetic tests. All quantitative results are reported as mean ± standard deviation of n = 6 parallel specimens and analyzed by one-way ANOVA followed by Tukey HSD post-hoc test at α = 0.05. The optimized harness scheme reduces the maximum interfacial shear stress from 18.46 ± 0.42 MPa to 10.58 ± 0.31 MPa (42.7 % reduction, p < 0.001), while the experimental average peeling force reaches 246.3 ± 4.8 N at 23 °C—23.1 % above the 200 N industrial threshold. After 1200 cycles of −30 °C↔+80 °C thermal cycling, 1000 h of 85 °C / 85 % RH hygrothermal exposure and 500 h of UV irradiation, no interlayer delamination or open-circuit failure is observed. Continuous transmittance is tunable within 12.3–77.8 %, full-band (470–860 MHz) DTV antenna gain fluctuation is bounded to ±1.3 dB (shielding effectiveness 11.4 dB at 470 MHz), and the dual-variant mass-production qualification rate over 42,716 inspected parts improves from 76.3 % to 99.2 %. To the authors’ knowledge, this work represents the first integrated thermo-mechano-electromagnetic numerical framework for triple-functional laminated windshield composite interfaces, supported by a self-developed unified-datum coupon-grade verification platform.

Computer Virus: An Analysis on Detection and Prevention Technology

Haradhan Kumar Mohajan — 2026-08-06

A computer virus is a destructive computer program that can spread across computers and networks by making copies of itself to a host file or system area commonly through the email attachments, usually without the users’ knowledge and runs against wishes of users. It attaches itself to files, programs or parts of the operating system and performs harmful actions, such as delete of files, slowdown of the computer, corruption of data, steal of sensitive information, damaging computers and computer systems, etc., of individuals and organizations. It is the most potent and vulnerable threat for computer users that hampers important work involved with data and documents. It has replication, protection, and payload mechanism. It is potentially dangerous that can threaten the proper functioning of the computer system. Once it gets control, it multiplies itself to form newer generations. An attempt has been taken here to discuss the detection and prevention strategies of computer viruses.