Search Results (3,414)

Rolling bearings exhibit nonlinear and non-stationary fault signals under complex working conditions, rendering single-modal representation insufficient for accurate diagnosis. To address this limitation, this paper proposes a novel parallel multimodal fusion fault diagnosis model based on a Gated Recurrent Unit (GRU), a Residual Network (ResNet), and a Convolutional Block Attention Module (CBAM). First, a systematic multimodal representation selection framework is introduced, identifying the Markov Transition Field (MTF) as the optimal two-dimensional (2D) image modality due to its superior texture clarity and noise resistance compared to other methods. Second, parallel dual-branch architecture is designed to simultaneously process heterogeneous data. The 1D-GRU branch captures long-range temporal dependencies directly from raw vibration signals, while the 2D ResNet-CBAM branch extracts deep spatial features from the MTF images, adaptively focusing on key fault regions. These heterogeneous features are then fused through concatenation to retain complementary diagnostic information. Experimental validation on the Case Western Reserve University (CWRU) dataset demonstrates that the proposed model achieves a 99.57% accuracy in a 10-classification task. Furthermore, it exhibits significant parameter efficiency and outstanding robustness, with the accuracy decreasing by no more than 1.2% under noise interference and cross-load scenarios, comprehensively outperforming existing single-modal and advanced fusion methods. Full article

Monocular 3D reconstruction remains a persistent challenge for autonomous driving systems in Degraded Visual Environments (DVEs) with extreme glare and low illumination, such as highway tunnels, due to the lack of reliable texture cues. This paper proposes a physics-aware deep learning framework that overcomes these limitations by fusing polarization sensing with conventional intensity imaging. Unlike traditional end-to-end data-driven fusion strategies, we propose a Modality-Aligned Parameter Injectionstrategy. By remapping the weight space of the input layer, this strategy achieves a smooth transfer of the pre-trained Vision Transformer (i.e., MiDaS) to multi-modal inputs. Its core advantage lies in the seamless integration of four-channel polarization geometric information while fully preserving the pre-trained semantic representation capabilities of the backbone network, thereby avoiding the overfitting risk associated with training from scratch on small-sample data. Furthermore, we design a Reliability-Aware Gating mechanism that dynamically re-weights appearance and geometric cues based on intensity saturation and the physical validity of polarization signals as measured by the Degree of Linear Polarization (DoLP). We validate the proposed method on our self-constructed POLAR-GLV benchmark, a real-world dataset collected specifically for high dynamic range tunnel scenarios. Extensive experiments demonstrate that our method consistently outperforms intensity-only baselines, reducing geometric reconstruction error by 24.2% in high-glare tunnel exit zones and 10.0% at tunnel entrances. Crucially, compared to multi-stream fusion architectures, these performance gains come with negligible additional computational cost, making the framework highly suitable for resource-constrained onboard inference environments. Full article

This study aims to determine the effects of incorporating melon and watermelon kernel oils into ice cream formulations on the textural profile, mineral richness, and antioxidant activity of the product, and to investigate how oil addition optimizes critical quality parameters such as melting characteristics and viscosity of ice cream. The parameters analyzed include dry matter percentage, first drop, meltdown, overrun, antioxidant content, color and textural characteristics, total phenolic content, and mineral matter content. Among the samples, the highest first drop, meltdown, and overrun values were determined to be 31.67 s, 122.08 s, and 33.34%, respectively, in ice cream samples produced with 0.3% melon kernel oil addition, and the highest DPPH, ABTS, FRAP, and TPC in samples produced with a 0.3% addition of watermelon kernel oil, with values of 81.88%, 9.90 µmol TE/g, 2.06 µmol TE/g, and 128.72 mg GAE/100 g, respectively. Likewise, the lowest firmness, highest consistency, cohesiveness, and viscosity index values (15.53 g, 456.34 g.s, −21.50 g.s, and −8.16) were also found in the same ice cream samples. P, Mg, Ca, Na, K, Fe, and Zn contents increased with increasing addition of seed oil, and P showed the highest increase among the samples, followed by Na, K, and Ca, respectively. The samples demonstrating the most significant increase in mineral content were those produced with 0.3% melon kernel oil. Full article

Detecting ships using Synthetic Aperture Radar (SAR) in coastal areas is still difficult due to the impact of coherent speckle noise from the ocean surface, complex land clutter and having multi-scale target representations in the radar imagery. Most of the existing ship detection algorithms lose important target features during downsampling and have difficulty recovering those features through upsampling, resulting in a high number of false detections and missed detections. In this work, we present a new ship detection algorithm called Synergistic Star-Attentive Network with Dynamic Feature Refinement (S²D-Net). First, we create a new backbone called Multi-scale PCCA-StarNet to generate robust feature representations. Within the backbone we implement a Progressive Channel-Coordinate Attention (PCCA) mechanism to create a synergy between global channel filtering and adaptive coordinate locking to decouple ship textures from granular speckle noise. Second, we create a Dynamic Feature Refinement Neck. We develop a content-aware dynamic upsampler called DySample to replace conventional interpolation to improve fidelity of the upsampled feature of small targets. Further, we design a Star-PCCA Feature Aggregation module which fuses features together. Using star-operations and the PCCA mechanism, this module refines semantic features and removes background clutter while aggregating features across multiple scales. Third, we develop a Lightweight Shared Convolutional Detection Head with Quality Estimation (LSCD-LQE). The LSCD-LQE decreases parameter redundancy by using shared convolutional layers and adds a localization quality estimation branch. Therefore, the LSCD-LQE effectively reduces false positive detections through alignment of classification scores with localization quality based on Intersection over Union (IoU) in difficult coastal environments. Our experimental results, using the SSDD and HRSID datasets, show that S²D-Net produces results comparable to representative ship detection algorithms. In particular, on the challenging HRSID inshore subset, our proposed method achieved a mean average precision (mAP) of 82.7%, which is 6.9% greater than the YOLOv11n baseline ship detection algorithm. These results demonstrate that S²D-Net is superior at detecting small coastal vessels and mitigating the detrimental effects of the nearshore complex environment on the performance ship detection using SAR. Full article

For structural metallic materials, performance enhancement has traditionally relied on complex adjustments of chemical composition and heat treatment processes. However, these approaches are complex, costly, and lack sustainability. Metal additive manufacturing (AM) has unique cooling characteristics, providing it with a distinctive approach. In this study, laser powder bed fusion (LPBF) technology was used to prepare high-performance 1080 carbon steel. The study selected three groups of process parameters (VED = 92.59 J/mm³) with high density (relative density > 98%) and achieved excellent mechanical properties: the ultimate tensile strength (UTS), yield strength (YS), and elongation (EL) reach 1745.4 MPa, 1455.13 MPa, and 6.77% respectively. The effects of process parameters on microstructure and mechanical properties were investigated. It is found all specimens exhibited a characteristic martensitic needle-like grain morphology without significant crystallographic texture. The microstructure displayed substantial changes as VED varied, with martensite content progressively decreasing with increasing VED. Correspondingly, as the VED increases from 92.59 J/mm³ to 225.69 J/mm³, the UTS, YS, and EL decrease by 39.0%, 36.1%, and 3.4%, respectively. This work demonstrates the feasibility of achieving high-performance metallic components by precisely controlling additive manufacturing process parameters to manipulate the microstructure of simple alloys, thereby eliminating the need for complex alloying or post-processing heat treatments. Full article

In this study, we investigated hyaluronic acid (HA) as a functional biopolymer for improving the processing performance of myofibrillar protein (MP) gels. Our focus was on the order of incorporation and concentration of HA as controllable process parameters, and their effects on water-holding capacity, rheological behaviour, texture, colour and microstructure of MP gels. The experimental results demonstrated that HA promoted the formation of a denser and more homogeneous protein network, as confirmed by microstructural analysis and significantly enhanced water retention. From a mechanical perspective, HA incorporation decreased hardness and chewiness while increasing adhesiveness, thereby improving overall gel functionality. Importantly, the simultaneous dissolution of HA with meat and water produced superior outcomes compared to post-addition, highlighting the role of ingredient addition sequence as a relevant process design factor. The slight colour variations remained within acceptable quality limits. Our findings provide new insights into protein hydrocolloid interactions in gel systems and indicate how HA can be strategically integrated into processing operations to improve product yield, quality and consumer acceptance in the meat industry. Full article

In the field of intelligent inspection, high-definition video data collected by quadruped robot dogs face severe transmission and storage constraints. Although existing advanced lossy video coding standards can significantly improve compression efficiency, they inevitably introduce severe compression artifacts in low-bit-rate scenarios. To address this issue, this paper proposes a video decoding quality enhancement network named Video Quality Restoration Network (VQRNet), based on a dual-stream architecture. Specifically, the Local Feature Extraction component incorporates a Progressive Feature Fusion Module (PFFM) with a four-stage progressive structure. By integrating reparameterized convolution and attention mechanisms, PFFM focuses on capturing high-frequency texture details to repair small-scale distortions. Simultaneously, the Multi-Scale Lightweight Spatial Attention Module (MLSA) performs spatial feature recalibration, leveraging multi-scale convolution to adaptively identify and enhance key spatial regions, specifically addressing multi-scale distortion. In the Global Feature Extraction component, the State-Space Attention Module (SSAM) combines State-Space Models (SSMs) with attention mechanisms to capture long-range dependencies and contextual information, for large-scale distortions caused by high-intensity compression. To verify the performance of the proposed algorithm, a dedicated dataset comprising 20 real-world video sequences captured by quadruped robot dogs (partitioned into 15 training and 5 testing sequences) was constructed, and the VTM 23.4 reference software was employed to simulate compression degradation using four quantization parameters (QP 30, 35, 40, and 45). Experimental results demonstrate that VQRNet outperforms state-of-the-art quality enhancement methods in terms of core metrics, including PSNR and SSIM, specifically including MIRNet, NAFNet, TRRHA, and CTNet. In the QP = 30 scenario, VQRNet achieves an average PSNR of 40.33 dB, a significant improvement of 3.32 dB over the VTM 23.4 baseline (37.01 dB), while demonstrating significant advantages in computational complexity and parameter efficiency—requiring only 5.27 G FLOPs and 1.40 M parameters, with an average inference latency of only 11.82 ms per 128 × 128 patch. This work provides robust technical support for the efficient video perception of quadruped robot dogs. Full article

Reducing cake fat while maintaining aeration, crumb softness, and consumer acceptance remains challenging because fat crystals contribute to interfacial stabilization and structure development. This study evaluated an interfacial processing strategy in which oil dispersion is refined by pre-emulsification to evaluate whether refining oil dispersion by pre-emulsification modulates the functional impact of lipase (via in situ formation of surface-active lipolysis products). A D-optimal design (16 formulations) quantified the effects of fat type (shortening vs. sunflower oil), fat level (100% vs. 50%), pre-emulsification (absent/present), and lipase dose (0, 50, 100 ppm; flour basis) on batter and baked-cake quality. Responses included moisture, color, volume/visual structure, texture and hedonic sensory evaluation for selected formulations. Lipase improved structure and texture, with the strongest benefits in reduced-fat samples, where hardness-related parameters decreased and volume/crumb refinement improved. Pre-emulsification modulated lipase performance in a formulation-dependent manner, indicating significant interactions. In sensory tests, the combined approach improved low-fat acceptance compared with the low-fat control. Overall, pre-emulsification-enabled lipase action offers a route to recover key quality attributes in low-fat cakes without conventional emulsifiers. Full article

Thin, dry skin is characterized by impaired barrier integrity, loss of dermal density, and accelerated aging driven by intrinsic and extrinsic factors. Biomimetic collagen peptides mimic native collagen sequences, stimulating fibroblasts to enhance synthesis while limiting matrix metalloproteinase-mediated degradation. This study evaluated the clinical efficacy and safety of a multi-ingredient cosmetic product for thin, dry, aging skin, formulated as a dual-purpose body and face serum lotion containing 0.1% biomimetic collagen tripeptide (Tripeptide-29) along with Niacinamide, Citrullus lanatus fruit extract, and Selaginella lepidophylla extract. In this prospective, single-center study, 47 healthy women, aged 36–65 years with Fitzpatrick skin types I–IV, applied the formula twice daily to the face and body over four weeks. Objective measurements—including elasticity, wrinkle depth and volume, hydration, trans-epidermal water loss (TEWL), and texture—were collected weekly alongside clinical grading and self-assessments. Significant improvements were observed across all parameters, with facial dryness decreasing immediately (−74.6%) and continuing to week 4 (−93.7%), hydration increasing up to 72.5%, softness improving up to 37.7%, roughness decreasing up to 37.9%, and TEWL reductions indicating strengthened barrier function. Desquamation improved by 75.5% by week 3, and no adverse effects occurred. The serum lotion demonstrated robust, well-tolerated benefits for enhancing multiple markers of thin, dry, aging skin. Full article

Although processing windows have been widely reported for LPBF Ti-6Al-4V, the distinct roles of laser power, scanning speed, and hatch distance remain unclear beyond VED-based comparisons. In this work, the distinct effects of laser power, scanning speed, and hatch distance on the microstructural evolution and mechanical response of laser powder bed fusion (LPBF) Ti-6Al-4V (Ti64) are investigated within a stable processing window with comparisons among different parameter combinations at a comparable VED. A total of 56 processing conditions were designed, and microstructure/texture and properties were characterized by OM/SEM, EBSD, microhardness (HV0.5), and hole-drilling residual stress measurements. Within the selected processing window, prior-β grain morphology, α’ martensite thickness, texture, microhardness, and residual stress exhibit distinct sensitivities to different processing parameters. Specifically, lower scanning speeds and smaller hatch distances promote more continuous <001>β epitaxial growth, whereas higher scanning speeds or larger hatch distances produce fragmented prior-β grains. The α’ lath thickness shows the strongest dependence on scanning speed with a secondary influence from hatch distance, while laser power mainly provides an overall thermal modulation. Furthermore, the macroscopic α (0002) texture is mainly governed by the β solidification texture, with α-variant selection playing a secondary, amplifying role. In addition, microhardness correlates with α’ martensite thickness following a Hall–Petch equation. The peak residual stress is more sensitive to scanning speed, while bulk residual stress varies more significantly with hatch distance. These findings demonstrate that process parameters, in addition to VED, can guide microstructural control and mechanical optimization in LPBF Ti64 alloy. Full article

In the study, the effects of adding sour cheery seed powder at different concentrations (0%, 0.5%, 1%, and 1.5%) on the pH, water content, lipid oxidation, cooking loss, color, TPC, antioxidant activity (DPPH and ABTS), texture, and acrylamide contents of meatballs cooked at 150 °C, 200 °C, and 250 °C were investigated. The sour cherry seed powder significantly affected the pH, cooking loss, a*, b*, C*, h°, TBARS, acrylamide, hardness, and springiness, while no significant effect was found on the moisture, L*, cohesiveness, gumminess, or chewiness. The cooking temperature had a significant effect on all the parameters except cohesiveness, gumminess and chewiness. The addition of 1% sour cherry seed powder resulted in the lowest acrylamide and TBARS values. The sour cherry seed powder increased the total phenolic content (TPC) and antioxidant activity of the meatballs. These results indicate that sour cherry seed powder can be used as a sustainable food ingredient in meatball production, reducing acrylamide and lipid oxidation while improving the antioxidant capacity, color, and texture properties. Full article

We studied the shelf life of fresh buffalo meat in polyamide/polyethylene (PA/PE) packaging during refrigerated storage for 14 days, when treated with cinnamon–clove (C-C) and nutmeg (Nut) powders, along with lavender essential oil (LEO). Microbiological (total viable count, Pseudomonas spp., Brochothrix thermosphacta, Enterobacteriaceae, and lactic acid bacteria), antibacterial (Salmonella Typhimurium and Staphylococcus aureus), physicochemical and biochemical (pH, moisture, color, total fat, hemoglobin and heme iron, 2-thiobarbituric acid, mercaptans, antioxidant activity, and total phenolic content), and sensory (color, odor, texture, and taste) analyses were carried out. The results showed that C-C and Nut powder extracts exhibited significant (p < 0.05) antioxidant and antibacterial activity, higher than LEO; however, all treatments delayed lipid oxidation. Based primarily on sensory evaluation, the shelf life extension of buffalo meat was 2–3 days for LEO and Nut powder, and 4–6 days for C-C powder. Factor analysis indicated the critical days of refrigerated storage for the evolution of spoilage-related biochemical parameters. Full article

This study was conducted to investigate and identify correlations among sensory and comprehensive consumer test results with rheological, textural, and tribological properties of milk chocolate in response to varying levels of particle size and emulsifier. To simulate realistic oral conditions, artificial saliva was incorporated into instrumental analyses. Rheological analysis revealed that increasing particle size and emulsifier concentration significantly reduced plastic viscosity, while emulsifier concentration alone increased yield stress due to structural reorganization within the fat phase. Tribological measurements demonstrated that larger particles increased friction in boundary and mixed lubrication regimes, whereas emulsifiers reduced friction in these regimes by enhancing fluid film formation. Under elastohydrodynamic conditions and with artificial saliva, friction was more influenced by the interaction between particle size and emulsifier level. Textural analysis showed that both parameters significantly influenced hardness, with saliva further softening the samples, especially those with higher emulsifier levels. Sensory evaluations indicated that emulsifiers enhanced flavor release and mouthfeel attributes, while smaller particles contributed to smoother texture and more balanced flavor perception. Consumer acceptance tests confirmed that samples with smaller particles and higher emulsifier levels received the highest scores in overall liking, taste, and texture. Instrumental parameters strongly correlated with key sensory attributes, with yield stress showing the highest positive associations with creaminess, smoothness, fat/milk flavor, and liking, while higher viscosity and friction were negatively linked to flavor release and mouthfeel. Instrumental hardness negatively correlated with cacao intensity and astringency, while saliva-induced softening was positively associated with sweetness and liking, highlighting the role of dynamic oral softening. Full article

Deep-sea gear transmission systems encounter critical lubrication challenges arising from the synergistic coupling of extreme hydrostatic pressure and cryogenic temperatures. These environmental stressors induce exponential viscosity escalation in lubricants, precipitating severe fluidity degradation, elevated startup resistance, and lubrication starvation. Concurrently, seawater intrusion triggers lubricant emulsification, additive deactivation, and electrochemical corrosion at meshing interfaces, collectively escalating the risk of catastrophic lubrication failure and compromising long-term operational reliability. This study systematically elucidates the lubrication degradation mechanisms inherent to deep-sea environments and proposes targeted mitigation strategies. Through comprehensive characterization of deep-sea environmental parameters and their impact on lubricant rheological behavior, we critically evaluate the applicability and inherent limitations of conventional Thermal Elasto-Hydrodynamic Lubrication (TEHL) theory under extreme conditions. Our analysis reveals that established TEHL frameworks necessitate substantial modification to accurately capture pressure-viscosity-temperature coupling phenomena and seawater contamination kinetics. Meshing interface texturing, as an effective anti-friction and wear-mitigation strategy, is investigated to delineate its mechanistic pathways for enhancing lubricant film formation and tribological performance under starved lubrication regimes. Key findings demonstrate that optimized micro-texture architectures can effectively compensate for viscosity-induced fluidity deficits and attenuate the deleterious effects of seawater ingress. Critical knowledge gaps are identified, and future research trajectories are charted: (i) multiphysics coupling models integrating thermo-hydrodynamic, chemo-physical, and mechanical degradation processes; (ii) synergistic texture-coating design paradigms; (iii) high-pressure low-temperature experimental validation protocols; and (iv) engineering implementation frameworks for deep-sea gear transmission systems. This review establishes theoretical foundations and provides technical guidelines for robust lubrication design and long-term operational stability of deep-sea transmission equipment. Full article

Using the ultrasonic burnishing process to fabricate micro-textures is one of the effective methods to improve the hydrophobic properties of workpiece surfaces. In this study, three ultrasonic burnishing strategies—single-pass ultrasonic burnishing process (SUBP), two-pass ultrasonic burnishing process with reverse feed direction (TUBP-RF), and two-pass ultrasonic burnishing process with forward feed direction (TUBP-FF)—were employed to fabricate micro-textures on 316 stainless steel pipes. The effects of burnishing strategy and feed rate on surface morphology and hydrophobic performance were investigated. TUBP-RF introduces reverse feed in the second pass, generating tangential forces in the opposite direction that induce secondary plastic flow and material accumulation at texture intersections. The results show that surface hydrophobicity first increased and then decreased with increasing feed rate, reaching its maximum at 0.7 mm/r. TUBP-RF achieved the highest contact angle of 108°, representing increases of 18.4% and 12.1% compared with SUBP and TUBP-FF, respectively. Among the three strategies, TUBP-RF produces interlaced micro-textures with larger peak height Rp, medium peak spacing RSm, and reduced effective solid contact area, facilitating air entrapment beneath water droplets and promoting a Cassie–Baxter wetting state. Furthermore, under the optimal parameters of the TUBP-RF process, the machined surface improved droplet sliding speed, reduced the sliding time by 61.7% compared with the original surface. The TUBP-RF strategy effectively enhances surface hydrophobic properties by constructing interlaced micro-textures, offering new insights for optimizing the ultrasonic burnishing process. Full article