Easwari Engineering College

In recent days, a substantial amount of research work has focused on speech improvement. A speech enhancement system is used in numerous applications like voice activity detection, human-machine interaction systems and acoustic emotion identification. Estimating noise statistics is a challenging task in the speech enhancement process while using non-stationary real-time models. Hence, a novel filtering-based speech enhancement framework is introduced in this work. The speech enhancement process is initiated by collecting the speech signal from standard resources. Further, the gathered signal is then processed for enhancement. Here, the new technique named Hybrid Filtering is introduced with the combination of Non-Local Means (NLM) and Kalman Filtering (KF), and this NLM approach performs the denoising operation on the collected speech signal. The KF effectively estimates the noise from the speech signal. Further, the speech enhancement performance of the Adaptive NLM with Kalman Filtering (ANLM-KF) is enhanced by optimizing the parameters with Innovated Clouded Leopard Optimization (ICLO). The proposed model efficiently enhances the speech signal by performing a Multi-scale denoising operation. Finally, the experimentation is conducted on the developed model over traditional methodologies to prove its effectiveness in the speech enhancement process. The proposed ICLO-ANLM-KF algorithm significantly outperforms all baseline optimization models across every quality metric. It achieves the lowest error rates, with a Mean Absolute Error (MAE) of 0.047 dB and a Root Mean Square Error (RMSE) of 0.216 dB. Moreover, the developed method yields the highest signal reconstruction quality, with a Peak Signal-to-Noise Ratio (PSNR) of 61.441 dB, which is a substantial improvement over models like DE-MCD, STC-STWF, SS-DWT, and TCN-MHA-Bi-GRU. These quantitative results confirm that the proposed ICLO-ANLM-KF is the most effective approach for enhancing speech signals while minimizing distortion.

Abstract This review critically examines the impact of carbonaceous nanofillers and fiber hybridization on the mechanical, thermal, and electrical characteristics of polymer composites. Advanced nanofillers such as graphene, carbon nanotubes, and carbon black are assessed for their potential to improve strength, conductivity, and thermal stability. Micro and macro-scale fibers (including carbon, glass, Kevlar, and natural fibers) are examined in hybrid configurations with thermoplastic and thermoset polymer matrices. Literature published between 2000 and 2025 was identified through a structured literature search of Scopus, Web of Science, and Google Scholar and synthesized to elucidate comparative performance trends. The review highlights dispersion strategies, interfacial engineering, and surface modification techniques that influence filler-matrix and fiber-matrix interactions, particularly emphasizing synergistic reinforcement mechanisms such as multi-scale interactions via crack-bridging by fibers and crack-deflection by nanofillers, along with the essential function of the interphase region in effective load transfer. Comparative analyses across reported studies demonstrate substantial improvements in tensile strength, flexural modulus, and thermal and electrical conductivity, often exceeding predictions based on the rule of mixtures. However, challenges remain in controlling nanofiller dispersion, maintaining strong interfacial compatibility, achieving effective hybridization, balancing multifunctional properties without performance trade-offs, ensuring scalable processing, and improving recyclability. The review integrates mechanistic insights and design methods to enlighten the development of high-performance, multifunctional polymer composites for aerospace, automotive, electronics, and construction applications.

The increasing discharge of synthetic dyes and pathogenic microorganisms into aquatic environments presents significant environmental and public health challenges, necessitating the development of sustainable and multifunctional treatment materials. In this study, an acid-functionalized biocarbon (AF-BC)-supported CeO2/Ga2O3 ternary nanocomposite was successfully developed as an efficient dual-functional material for photocatalytic wastewater treatment and antibacterial applications. The integration of Ga2O3 and conductive biocarbon markedly improved light absorption, promoted charge separation, and accelerated interfacial electron transport. The AF-BC/CeO2/Ga2O3 ternary composite demonstrated enhanced photocatalytic activity under natural sunlight, with 96

In this work, the tribological, mechanical, corrosion, and sliding wear behavior of cold metal transfer-wire arc additive manufactured (CMT-WAAM) Al4043 alloy is reported and optimized using the combined use of response surface methodology (RSM) and machine learning (ML). According to microstructural observations, the grain size at the bottom was 67 μm, finer than that at the top (93 μm) due to the higher cooling rate, which led to corresponding tensile strength and elongation. This fine-grained structure in the bottom region also contributed to higher hardness values (70.62 HV) and better wear resistance, as supported by the hardness profile and wear testing. Hardness was lower in the top quarter, which had larger grains. Tensile testing demonstrated that the 0° direction exhibits the highest UTS and elongation due to the fine microstructure developed at an optimal cooling rate. Corrosion test. The corrosion rate of the top part was the lowest (0.098 mm/year) and showed the greatest potential for corrosion protection (Ecorr: − 751.23 mV), indicating a stable passive oxide film. Optimum conditions (33.3 N load, 362 RPM speed, and 50 mm wear track radius) derived by RSM optimization minimized the specific wear rate (SWR) and coefficient of friction (COF), thereby improving wear response. The ML model of XGBoost correctly predicted SWR and COF with an excellent fit in terms of R2 value (0.992 and 0.998, respectively), verifying the validity of the model. This work systematically enlightens the optimization of tribological, mechanical, and corrosion behaviors of WAAM Al4043 alloys, shedding light on their applications in engineering.

The continuous increase in power density and packaging density of electronic devices has imposed stringent requirements on thermal interface materials (TIMs) to deliver high thermal conductivity, electrical insulation, processability, and long-term reliability. In this study, an hBN-rich hybrid TIM is developed by integrating graphene oxide (GO), zinc oxide (ZnO), silicon carbide (SiC), and hexagonal boron nitride (hBN) within a silicone matrix without employing silane-based surface modification. The formulation strategy relies on multi-dimensional filler synergy to construct efficient heat-conduction networks while preserving dielectric integrity. The optimized hybrid composition (12 wt