The SAW (surface acoustic wave) gyroscopic effect is a key parameter that reflects the sensitivity performance of SAW angular velocity sensors. This study found that adding a layer of non-piezoelectric material with a lower reflection coefficient than that of the upper-layer material below the piezoelectric substrate to form a double-layer structure significantly enhanced the SAW gyroscopic effect, and the smaller the reflection coefficient of the lower-layer material, the stronger the SAW gyroscopic effect, with values being reached that were two to three times those with single-layer substrate structures. This was confirmed using a three-dimensional model, and the experimental results also showed that the thickness of the piezoelectric layer and the type of the lower-layer material also had a significant impact on the SAW gyroscopic effect. This novel discovery will pave the way for the future development of SAW angular velocity sensors.

A surface acoustic wave (SAW) gyroscope has many unique advantages, but a low detection sensitivity limits its development. Previous studies have shown that adding a metal dot array to the acoustic wave propagation path of the SAW delay line can enhance the Coriolis force and further improve sensitivity. Therefore, in order to optimize the detection sensitivity performance of the sensor, 128°YX-LiNbO₃, ST-X Quartz and X112°Y-LiTaO₃ piezoelectric substrates were selected by finite element method to analyze the influence of the metal dot array size on the SAW gyroscopic effect in this paper. The most suitable metal dot size for 128°YX-LiNbO₃ and X112°Y-LiTaO₃ obtained by simulation are $5/16\lambda$ and $1/16\lambda$, respectively; for example, when the normalized angular velocity is 1 × 10⁻³, the SAW gyroscopic effect factor $g$ of the two piezoelectric substrates distributing the optimum size metal dots can reach 22.4 kHz and 5.2 kHz. For ST-X quartz, there is a threshold between the rotation speed of the substrate and the optimum size of the metal dot. When the rotating speed is lower than the threshold, the SAW gyroscopic effect is strongest when the metal dot size is $3/16\lambda$; otherwise, the SAW gyroscopic effect is strongest when the size is $11/16\lambda$. These research results provide new ideas for improvement of the SAW gyroscope.

Plasma etching is a critical step in semiconductor manufacturing, yet existing approaches are either computationally expensive or limited to predicting scalar etching metrics rather than full profile evolution. We propose EPreNet, a condition-guided spatio-temporal network for pixel-level prediction of plasma etching profile evolution from historical profile frames and process parameters. To support this task, we construct a benchmark dataset of 18,360 images spanning 918 process conditions simulated via TCAD, sampled with Latin Hypercube Sampling to ensure uniform parameter-space coverage, and further establish an evaluation framework combining image-level and geometry-based metrics for etching-profile prediction. Experiments demonstrate that EPreNet reduces MSE by 16% and achieves SSIM of 0.992 and PSNR of 30.323 dB, while achieving manufacturing-relevant geometric accuracy with $1.4°$ sidewall angle error and 1.6% depth error rate. Inference requires only 38.92 ms per frame faster than TCAD simulation 1300 s, supporting rapid surrogate-based evaluation and accelerated TCAD-assisted process exploration. The model also shows strong generalization to unseen initial critical dimensions and encouraging initial transferability to preprocessed experimental SEM images, suggesting its potential as an efficient surrogate for TCAD-assisted process development while maintaining high geometric fidelity.

Many high-performing camouflaged object detection (COD) methods rely on large models and substantial computational costs, limiting their deployment on resource-constrained devices. Lightweight backbones reduce model size, but their limited representational capacity makes it challenging to distinguish weak object structures from background-similar textures. Frequency-domain and boundary-guided methods provide complementary structural cues, making their selective use important for lightweight COD. However, both background textures and object contours may produce strong high-frequency responses, while propagating erroneous edges can amplify background interference. We present AmbiCOD, a lightweight network that conditions structural cues on semantic context. The Frequency-Semantic Discrepancy Modulation (FSDM) module jointly exploits shallow frequency responses, deep semantic responses, and their spatial discrepancy to adaptively enhance or attenuate shallow structural features through bounded bidirectional residual modulation. The Semantic-Structure Calibration Module (SSCM) coordinates multiscale representations, and the Reliability-Gated Refinement Decoder (RGRD) combines boundary probabilities with task-conditioned modulation weights to guide progressive mask reconstruction. With 3.00 M parameters and 4.16 G FLOPs, proposed AmbiCOD achieves the highest S_{\alpha } among the compared lightweight methods on all four COD benchmarks. Code is available at https://github.com/AaHa123/AmbiCOD.git.