Search Results (356)

The sustainable production of leafy vegetables in Mediterranean and East Asian regions is increasingly constrained by water scarcity and nutrient imbalance in soil–plant systems, making timely and accurate nutrient diagnosis essential for precision fertilization. Conventional tissue analysis of nitrogen (N), phosphorus (P), and potassium (K) in lettuce is destructive, costly, and time-consuming, while existing non-destructive approaches based on traditional machine learning or deep learning still suffer from limited accuracy and poor generalization. To address these limitations, this study proposes ResNet-SA, a residual convolutional network assisted by a multi-head self-attention (MSA) mechanism, for the rapid and non-destructive estimation of leaf N, P, and K contents from top-view RGB images of lettuce trays under soilless cultivation. Two fusion strategies between the MSA mechanism and the ResNet50 trunk were evaluated, namely lateral side-connection of the attention block (ResNet50_SA_R series) and replacement of a trunk stage (ResNet50_SA_E series), each with three insertion depths. In the fixed-split evaluation, ResNet50_SA_RV1 achieved the best performance, with a test-set R² of 0.92 versus 0.81 for the ResNet50 baseline (absolute R² gains of 0.11 and 0.08 for RV1 and RV3, respectively). Grouped five-fold cross-validation by sampling batch tentatively verified the stability of this improvement (ResNet50_SA_RV1: R² = 0.92 ± 0.03; RMSE = 7.16 ± 1.77 mg/g; MAE = 4.06 ± 1.58 mg/g, macro-averaged across N, P, and K), with significantly lower prediction error than both the ResNet50 baseline (ΔMAE = −2.25 mg/g, 95% CI: −2.88 to −1.62, Holm-adjusted p < 0.001) and four conventional CNN architectures (R² range: 0.49–0.80). These results demonstrate that integrating the MSA mechanism into ResNet provides a reliable, non-destructive tool for lettuce nutrient diagnosis, offering practical support for precision fertilization and sustainable greenhouse production. Full article

Enhanced circular tubes are widely employed in shell-and-tube heat exchangers, power-generation condensers, chemical reactors, refrigeration systems, and air-cooled heat exchangers, where improved convective performance can reduce the heat-transfer area required for a specified thermal duty. Helical screw tapes (HSTs) are passive inserts that promote sustained swirling motion and enhance convective heat transfer within such tubes. Although helical screw tapes and multiple-insert arrangements have been investigated previously, the combined thermohydraulic and second-law effects of increasing the number of co-rotating HSTs under fixed geometric ratios remain insufficiently quantified. In this investigation, turbulent airflow in a heated round tube was numerically investigated to examine the effect of tape number on heat transfer, pressure drop, thermal performance, total entropy generation (S_total), and exergy destruction (Ex_D). Six HST configurations containing one to six tapes were examined over a Reynolds-number range of Re = 5000–20,000 in a circular tube with an inner diameter of D_T = 31 mm, which was also adopted as the characteristic length for the Reynolds number, Nusselt number, and friction factor. The helical pitch P, screw diameter D_s, tape width W, and tape thickness t were 60 mm, 30 mm, 4.5 mm, and 0.2 mm, respectively, giving a pitch ratio of P/D_s = 2.0 and a width ratio of W/D_s = 0.15. A plain tube (PT) served as the baseline case. The results show that increasing the number of tapes intensifies swirl flow and enhances heat transfer but also leads to a continuous increase in pressure loss. For the optimum three-tape arrangement, the Nusselt number is increased by 126.0–158.8% and the thermal performance factor by 4.5–19.5% relative to the plain tube, while the total entropy generation and exergy destruction are simultaneously reduced by 7.9–61.0%. Among the configurations examined, HST-P2.0-W0.150-3, comprising three tapes at a pitch ratio of P/D_s = 2.0 and a width ratio of W/D_s = 0.15, achieved the best overall performance by delivering the highest thermal performance factor and the lowest total entropy generation and exergy destruction among the HST cases. Full article

Peripheral neuropathy degrades gait mechanics, elevating peak plantar pressures (PPP) and tissue ulceration risks. This exploratory pilot study evaluates custom orthotics using thermo-mechanically synthesized re-entrant auxetic foam insoles with targeted dome-shaped inserts. Bilateral dynamic gait analysis was conducted across a heterogeneous cohort (N = 9) utilizing a P-Walk 600 pressure plate and a MARVUE 2D motion capture system inside standardized footwear. To address small-sample limits, a non-parametric bootstrap resampling analysis (B = 1000) was executed. Native auxetic foam demonstrated high internal consistency ((σ*) = 12.063 kPa, Consistency Rank = 1), showing a strong numerical propensity to restrict load distribution variability compared to factory-installed over-the-counter (OTC) memory foam controls. While initial cell-adaptation cycles were observed, custom insoles reduced PPP by up to 62.2% in systemic polyneuropathies, shifting signatures beneath the exploratory clinical safety target (<200 kPa) in 55% of simulated cases. Conversely, customization triggered an adverse volumetric crowding effect in focal mononeuropathies, suggesting un-customized native auxetic foam as the preferred structural configuration for this specific cohort to avoid premature cell densification. Re-entrant cellular structures accommodate heterogeneous pathomechanics, offering an exploratory pathway for patient-specific orthotic optimization. Full article

Main memory carries data outside the processor’s trust boundary, so commodity systems-on-chip (SoCs) increasingly encrypt it; yet, in-line memory encryption engine itself becomes a differential power analysis (DPA) target whose key, if recovered, unlocks all of dynamic random-access memory (DRAM). We present PACE, a page-adaptive, cache-anchored memory encryption engine for RISC-V that makes nth-order DPA resistance practical and keeps cryptographic latency off the cache eviction critical path. PACE inserts a TileLink adapter between the last-level cache and the memory port and applies, per physical page, one of four policies (plaintext/confidentiality/confidentiality+integrity/+masking-order-d) selected from RISC-V page table bits through a memory-mapped control plane. Confidentiality uses counter mode whose per-line keystream is precomputed during cache residency; integrity is tree-free at the embedded operating point via on-chip counters and tags, with a live split counter block-MAC Bonsai Merkle tree for scale-out. DPA resistance is layered: ISAP-style fresh re-keying caps the data complexity per key at $q\approx 1$, and domain-oriented masking (DOM, d + 1 shares) protects the sole key processing block to order d. We implement PACE in Chisel on a Rocket SoC (Chipyard) and evaluate it with open-source tooling. A deterministic TileLink-level harness proves ciphertext-in-memory and detects tamper/replay/splice, and the live Tier-B engine (DRAM counters and per-line message authentication codes (MACs) plus an on-chip-rooted block-MAC tree) is validated from end to end on full Rocket and BOOM SoCs and on the FPGA; the masked Ascon-p S-box is proven order-d secure (d = 1, 2) under a glitch- and transition-aware model by three independent formal tools (COCO, PROLEAD, and SILVER, the last also deciding the full composability lattice and confirming exact glitch-robust order-2 probing security), with COCO extending the exact verdict to the highest synthesized order d = 3 (secure at probing orders 1–3); a simulated trace correlation power analysis (CPA) recovers the full key from an unprotected core and is defeated by masking, with a mutual information analysis confirming the $N\propto {\sigma }^{2(d+1)}$ trace amplification law. We further realize PACE on field-programmable gate array (FPGA) silicon: the engine plus an on-chip ring oscillator power sensor is placed, routed, timing-closed at 100 MHz, and programmed on a Xilinx XC7Z020, and we drive a fixed-vs-random Test Vector Leakage Assessment (TVLA) campaign read back entirely over a JTAG (Joint Test Action Group). A multi-core configuration and a Linux control-plane driver are likewise validated. Across synthetic access patterns and named application kernels (AES, SHA-256, matrix multiplication, pointer chasing) on both in-order Rocket and out-of-order BOOM, application-level overhead is within measurement noise of plaintext for cache resident workloads (masking, in particular, is cycle-identical to plain confidentiality), and we characterize the cost of each policy, masking order, and re-keying interval, demonstrating side-channel-hardened memory encryption on open RISC-V hardware. Full article

The present study was performed to determine the antibiotic resistance genes (ARGs), virulence determinants, and mobile genetic elements in multidrug-resistant Escherichia coli CUK-76 isolated from wastewater in Himachal Pradesh, India. Whole genome sequencing was performed using the Illumina Miseq system, and the draft genome sequence was assembled by Unicycler v0.5.1 and annotated by the NCBI Prokaryotic Genome Annotation Pipeline (PGAP v6.10). The bioinformatics-based prediction analysis was performed using ResFinder v4.7.2 and CARD v4.0.1 (antibiotic resistance genes), VirulenceFinder v2.0, VFDB and MGEFinder v1.0.3 (virulence determinants), PlasmidFinder v2.0.1 (plasmid sequences), MLST v2.0 (sequence type), ISFinder and TnCentral v2.0 (insertion sequences and transposons), PathogenFinder2 v0.6.0 (pathogenicity), RAST (subsystems category) and Phigaro (prophage sequences). The draft genome of E. coli CUK-76 strain comprised 4,607,136 bp with a GC content of 51%. Genome annotation revealed 4498 genes of which 4289 were protein-coding genes, 78 RNA genes, and 131 pseudogenes. It was related to sequence type ST949 and its predicted resistome consisted of bla_CTX-M-15, bla_TEM-1B (class A β-lactamase genes), bla_EC-14 (class C β-lactamase gene), aph(6)-Id, aph(3″)-Ib, qnrS1, sul2, tet(A), and dfrA14 genes. Additionally, multiple virulence genes, two plasmid sequences viz. IncFIB(K) and IncFIB(AP001918), insertion sequences, transposons and prophage sequences were detected. The genomic dataset of this strain will be a valuable resource for comparative genomic studies on E. coli. Full article

This paper studies the Vehicle Routing Problem with Simultaneous Pickup and Delivery and Time Windows (VRPSPDTW). The problem arises from reverse logistics, last-mile distribution, and circular logistics, where vehicles must serve delivery and pickup demands while respecting vehicle capacity and customer time-window constraints. This work develops DBLS-SP. The method combines an improved insertion-based initialization procedure, a repair-oriented dynamic scoring strategy, ejection-based reinsertion, multi-armed-bandit control of the route-inheritance ratio, vehicle-layer archive-guided route descent, and quality-first nearest-neighbor pool replacement. Computational experiments on the Wang–Chen (WC) and JD Logistics (JD) benchmark instances show that DBLS-SP obtains the best algorithmic solution on 67 of 68 WC instances and the best or tied-best objective value on 16 of 20 JD instances. The ablation study confirms the contribution of the main components, and the proposed reduction strategy removes more infeasible paths than the original reduction rule while preserving feasibility. Full article

Zero-cost proxies reduce the cost of neural architecture search (NAS), the automated design of neural network architectures, by scoring untrained candidate architectures without full training. Although this makes individual evaluations inexpensive, zero-shot NAS still operates over a large architecture space and therefore depends not only on proxy quality but also on how proxy evaluations are allocated under a limited budget. Despite this role, many zero-shot NAS studies have primarily focused on proxy design, while search over large architecture spaces has often relied on standard evolutionary procedures with limited attention to fixed-proxy budget allocation. In this study, we propose Evaluation-Efficient Memetic NAS (EEM-NAS), a memetic framework for budgeted fixed-proxy zero-shot NAS. EEM-NAS combines bounded one-flip local search, rank-plus-distance target selection, refined-candidate reinsertion, and entropy-guided mutation within one fixed-proxy search process. These components exploit promising candidates, reduce redundant refinement, preserve useful local improvements, and maintain population-level exploration under the same proxy-evaluation budget. On NAS-Bench-201 TSS with eight zero-cost proxies, three datasets, and four fixed-proxy baselines under a budget of 100 proxy evaluations per run, EEM-NAS records 93 wins, three ties, and zero losses over the 96 proxy–dataset–baseline comparisons on final proxy score under paired statistical testing. Because a higher proxy score is informative of final accuracy only when the proxy ranks architectures well, we also report the macro-average test accuracy of the top-10 architectures found in each run, where EEM-NAS achieves the highest macro-average top-10 accuracy among the compared methods on the three most rank-aligned proxies. Full article

Cultural organizations, like museums, libraries, and archives, are trusted spaces for discovery, innovation, interrogation, and encouraging agency and exploration. They go beyond the physicality of objects, focusing on dissemination of knowledge as the very mission of the organization. By adopting this mission around using collection items, crowdsourcing projects, and programming to expose digital literacy concepts like bias, algorithms, and more, institutions have a new opportunity to become essential to life-long learning geared towards this second quarter of the 21st century. Through hands-on experiences that expose the bias and choice in all description and knowledge creation, these organizations can reinsert human-in-the-loop frameworks into machine learning and artificial intelligence platforms. Full article

In applications pertaining to sensor network, the Internet of Things, industrial monitoring, and cyber–physical security, graphs are increasingly being outsourced to clouds, where similarity search should be supported without exposing graph content, query graphs, update contents, or database evolution. Existing privacy-preserving graph similarity schemes mainly target static encrypted databases, and as such struggle to handle insertions, deletions, label updates, and long-running index maintenance. This paper proposes DFB-PPGSQ, a dynamic forward-private and epochal backward-private graph similarity matching scheme that moves branch-based lower-bound filtering into a structured encryption framework. DFB-PPGSQ uses epoch-local feature tokens, per-record occurrence handles, one-time update labels, update buffers, deletion tombstones, and shuffle-based branch–tree re-randomization to preserve pruning efficiency while making same-epoch tombstone, traversal, size, timing, and refresh leakage explicit. We formalize the system model, leakage functions, algorithms, and security interpretation, then implement a reproducible Python prototype with HMAC-SHA256 token generation and multi-profile dynamic sensor-topology workloads. Across five random seeds, DFB-PPGSQ keeps server-side filtering latency close to the static branch–tree baseline (46.60 ms versus 44.72 ms at 4000 graphs), avoids immediate full-rebuild updates (0.091 ms insertion and 0.092 ms label update), and keeps metadata-assisted cross-epoch token linkage below 5.6% attack success after refresh in additional industrial and campus IoT stress workloads. Storage, communication, exact GED refinement, side-channel hardening, and verifiable-result protection are treated as deployment costs and limitations rather than being included in the headline latency. Full article

Leaf disease diagnosis needs models that are accurate enough for agronomic use yet small enough for constrained computing settings. This study examines a late-attention MobileNetV2 design in which one Convolutional Block Attention Module (CBAM) is inserted between the last MobileNetV2 convolutional map and global average pooling. The experiments use 54,306 controlled-background PlantVillage images spanning 38 classes. Under a uniform saved-model re-evaluation, MobileNetV2 + CBAM obtained 97.17% accuracy and 97.15% weighted F1-score, whereas MobileNetV2 obtained 96.78% and 96.73%. On the converted models, paired testing gave a 0.64-percentage-point accuracy advantage for the CBAM variant (95% CI: 0.31–0.96; exact McNemar $p<0.001$). The proposed network has 4.02 million parameters, costs 0.604 GFLOPs (about 0.302 GMACs), and yields a 4.07 MiB dynamic-range-quantized TensorFlow Lite file with 96.70% accuracy. Batch-one inference on an Intel i7-11800H CPU with TensorFlow Lite/XNNPACK and eight threads reached a median of 45.42 ms (P95: 102.54 ms), excluding preprocessing. Grad-CAM inspection illustrates both lesion-centered activation and unresolved shared errors. The evidence therefore supports a compact accuracy–cost compromise for the tested conditions, while field robustness, energy use, repeated training runs, and target-device behavior remain open validation requirements. Full article

Background and Objectives: Vascular access in cancer patients with limited mobility remains a persistent clinical challenge, although both devices and insertion methods have improved considerably. We present our clinical experience with at-home PICC placement, supported by real-world data from routine practice, in oncology patients unable to attend hospital care, exploring its feasibility, safety, and practical applicability. Materials and Methods: All PICC insertions were performed at patients’ bedside by a trained surgeon using sterile technique. A single- or double-lumen PICC catheter was inserted under ultrasound guidance, magnetic tracking, and intracavitary ECG confirmation to secure optimal venous entry and accurate tip positioning. Results: A total of 28 PICC lines were successfully inserted in 26 oncologic patients. Two patients required catheter re-insertion, one following accidental removal and one due to infection. In 26 of 28 cases (92.9%), insertions were completed on the first attempt, while two cases required a second attempt. Insertion in the Green Zone per the ZIM classification was achieved in 26 of 28 cases (92.9%), and only two required insertions in the Yellow Zone. The basilic vein was chosen in 75% of patients. A total of 23 single-lumen and five double-lumen catheters were inserted. In terms of complications, four cases of systemic infection occurred more than 15 days after insertion and were not considered insertion-related; two patients developed localized edema requiring anticoagulation, and four catheter occlusions were managed conservatively. Conclusions: At-home PICC placement appears to be a feasible and clinically applicable approach when performed by qualified and credentialed professionals. These observations, derived from real-world clinical experience, suggest a potential role for this model in selected patient populations. Using proper sterile technique and real-time tip confirmation tools, this intervention may contribute to improved continuity of care and reduced hospital visits in selected patients. These observations suggest a potential role for expanding outpatient vascular access programs in selected patient populations. Full article

Background/Objectives: Maternal cell contamination (MCC) is a pervasive challenge in prenatal single-gene testing, as it can cause both false-positive and false-negative results. Despite its clinical significance, quantitative tolerance thresholds for MCC in whole exome sequencing (WES) and Sanger sequencing remain limited. Methods: We established a gradient contamination model (5–95%) using genomic DNA from 20 mother–child pairs and assessed the detection fidelity for single-nucleotide variants (SNVs) and insertions/deletions (InDels) in two clinically relevant scenarios using both WES and Sanger sequencing—Scenario 1 (Fetus Heterozygous/Mother Wild-type) and Scenario 2 (Fetus Wild-type/Mother Heterozygous). Results: In Scenario 1, analysis of the 20 loci evaluated by both WES and Sanger sequencing revealed false-negative rates of 0% when MCC ≤ 30%, which then increased sharply across the 30–70% MCC range and reached 100% at 95% MCC. In Scenario 2, analysis of 12,627 WES loci showed a false-positive rate of ≤1.0% when MCC ≤ 10%, which then increased from 1.0% to 76.6% within the 10–30% MCC range and climbed to 89.8% at 50% MCC. A similar pattern was observed for the 20 loci analyzed by both WES and Sanger sequencing in this scenario. These findings potentially support a three-tier stratification: low-risk (MCC ≤ 30% in Scenario 1 and ≤10% in Scenario 2), moderate-risk (30% < MCC < 70% in Scenario 1, 10% < MCC < 30% in Scenario 2), and high-risk (MCC ≥ 70% in Scenario 1 and ≥30% in Scenario 2). Conclusions: Our findings reveal a clear directional asymmetry of MCC tolerance between the two scenarios, indicating that MCC tolerance depends not only on contamination level but also on the genotype of the contaminating DNA. We further propose a clinical reference framework based on a three-tier risk stratification for prenatal single-gene testing, which can guide result interpretation and laboratory decision-making, including decisions about reliable reporting, orthogonal validation, or re-sampling. Full article

Phase resistance unbalance in robotic joint drive systems can alter electromagnetic torque generation and degrade motion accuracy, but its early diagnosis is challenging because fault-related signatures are weak and coupled with operating dynamics. This study proposes a short-time Fourier transform (STFT)-based multi-scale attention ResNet for phase resistance-unbalance diagnosis using synchronized multi-sensor signals from a single industrial robotic joint. Controlled resistance-unbalance states were generated on an eRob70F100I-BM-18EN joint module by inserting 0.05 $\Omega$ and 0.1 $\Omega$ series resistors into one motor phase with a nominal single-phase resistance of 0.75 $\Omega$. Current, acceleration, rotational speed, and torque signals were segmented and converted into four-channel STFT log-amplitude maps. A modified ResNet18 backbone was integrated with feature pyramid network (FPN)-style multi-scale fusion and a convolutional block attention module (CBAM) to enhance discriminative time–frequency features. Under a window-level stratified split, the proposed model achieved 98.97% accuracy and 98.97% macro-F1, outperforming raw-signal, fast Fourier transform (FFT), wavelet, STFT-ResNet18, STFT-VGG11-BN, STFT-MobileNetV2, and STFT-ShuffleNetV2 baselines. Grouped validation was conducted using file-level, leave-one-speed-out, and leave-one-load-out splits to assess robustness under stricter data partitions. The proposed model achieved 91.75% macro-F1 under file-level splitting and average macro-F1 values of 89.77% and 84.22% under leave-one-speed-out and leave-one-load-out validation, respectively. Grad-CAM visualization further indicates that the model relies on non-uniform local time–frequency regions rather than uniformly using the entire spectrogram. These results demonstrate effective robotic-joint resistance-unbalance discrimination while revealing that unseen operating conditions, especially specific speed and load settings, remain challenging for robust cross-condition deployment. Full article

Vibration measured at the cultivator shank is a rich but underutilized source of information about soil–tool interaction, which we propose can be read as a language of the implement’s engagement with the soil. As a first step of understanding this language, this study focuses on the vibration signature of the turning phase of the tillage operation. During a turn, the shank is lifted out of the soil, the turn is completed, and then reinserted into the soil. Using a convolutional neural network autoencoder, trained with a custom dual-domain loss function on primary tillage signals collected in 2022 on silt loam, a shared pattern was extracted with a mean Pearson correlation of 0.965. This pattern was then validated against independent 2023 primary tillage data, yielding strong agreement primarily in the frequency-energy distribution (FFT r = 0.93, PSD r = 0.99), while time-domain and time–frequency agreement were more moderate. When applied to a 2024 dataset from secondary tillage on Chase silty clay loam, the broad energy envelope was retained, but the primary-tillage time–frequency distribution was not reproduced. The proposed framework provides a practical foundation for future vibration-based interpretation of tillage operations without requiring traditional terramechanics assumptions. Full article

Kurgans in the Northern Black Sea region have long been treated as monuments “frozen” in their Scythian moment, yet excavation records show repeated re-engagement over much longer timescales. This article argues that kurgan reuse in the Western Scythian world is best understood not as a single behaviour, but as a bundle of practices with different archaeological signatures and social implications. Rather than offering an exhaustive catalogue, the study develops a methodological framework based on selected published case studies. It proposes a three-part typology: (1) intrusive mortuary reuse, that is, later secondary or inserted burials placed within earlier mounds; (2) associative reuse, including landscape appropriation, in which older barrows structured later funerary space through clustering, demarcation, visibility, and spatial reference; and (3) reuse through reopening, including shafts, tunnels, selective removal, and the manipulation of human remains. The article applies this framework to evidence from Scythian, Sarmatian, Greek, and early medieval/medieval contexts, with particular attention to Sarmatian intrusive burials in earlier barrows, the structured reuse of older kurgans in the Molochna basin, reopening and looting in Scythian and Sarmatian mounds, and the cross-cultural case of Koshary in the Olbian chora. The discussion situates these practices within mortuary archaeology, social memory, and the politics of the past, and argues that older kurgans could function not only as burial places, but also as mnemonic anchors, territorial markers, and targets of symbolic aggression. Full article