Current Issue : October-December Volume : 2025 Issue Number : 4 Articles : 5 Articles
We propose a deep learning-based face anti-spoofing method that utilizes both RGB and depth images for face recognition. The proposed method can detect spoofing attacks across various domain types using domain adversarial learning for preventing overfitting to a specific domain. A pre-trained face detection model and a face segmentation model are employed to detect a facial region from RGB images. The pixels outside the facial region in the corresponding depth image are replaced with the depth values of the nearest pixels in the facial region to minimize background influence. Subsequently, a network comprising convolutional layers and a self-attention layer extracts features from RGB and depth images separately, then fuses them to detect spoofing attacks. The proposed network is trained using domain adversarial learning to ensure robustness against various types of face spoofing attacks. The experiment results show that the proposed network reduces the average Attack Presentation Classification Error Rate (APCER) to 11.12% and 8.00% compared to ResNet and MobileNet, respectively....
In this paper, a novel through-silicon via (TSV) fabrication strategy based on through-hole structures is proposed for low-cost and low-complexity manufacturing. Compared to conventional TSV fabrication processes, this method significantly simplifies the process flow by employing double-sided liner deposition, double-sided barrier layer/seed layer formation, and double-sided Cu electroplating. This method enhances the TSV stability by eliminating Cu contamination issues during chemical–mechanical polishing (CMP), which are a common challenge in traditional blind via fabrication processes. Additionally, the liner and barrier layer/seed layer achieve a high step coverage exceeding 80%, ensuring excellent conformality and structural integrity. For electroplating, a multi-stage bidirectional electroplating technique is introduced to enable void-free Cu filling in TSVs. The fabricated TSVs exhibit an ultra-low leakage current of 135 fA at 20 V, demonstrating their potential for advancing 3D integration technologies in heterogeneous integration....
To shorten the development cycle of integrated circuit (IC) chips, third-party IP cores (3PIPs) are widely used in the design phase; however, these 3PIPs may be untrusted, creating potential vulnerabilities. Attackers may insert hardware Trojans (HTs) into 3PIPs, resulting in the leakage of critical information, alteration of circuit functions, or even physical damage to circuits. This has attracted considerable attention, leading to increased research efforts focusing on detection methods for HTs. This paper proposes a K-Hypergraph model construction methodology oriented towards the abstraction of HT characteristics, aiming at detecting HTs. This method employs the K-nearest neighbors (K-NN) algorithm to construct a hypergraph model of gate-level netlists based on the extracted features. To ensure data balance, the SMOTE algorithm is employed before constructing the K-Hypergraph model. Then, the K-Hypergraph model is trained, and the weights of the K-Hypergraph are updated to accomplish the classification task of distinguishing between Trojan nodes and normal nodes. The experimental results demonstrate that, when evaluating Trust-Hub benchmark performance indicators, the proposed method has average balanced accuracy of 91.18% in classifying Trojan nodes, with a true positive rate (TPR) of 92.12%....
In this work, a Ga2O3 self-switching nano-diode (SSND) fabricated on a sapphire substrate is presented. Unlike previous SSND studies that primarily focused on rectification, this work explores additional device characteristics, including breakdown voltage and photodetection functionality. The diode exhibits increased current with the increase in the relative permittivity (𝜿) of the device trenches, and under UV illumination. Devices with increased 𝜿 achieve a high current density of ≈10 kA cm−2, a rectification ratio of ≈104, and a specific on-resistance as low as 0.29 mΩ cm2. They also exhibit a breakdown voltage greater than 100 V, and a maximum power figure-of-merit of 35.46 MW cm−2. Moreover, under 240 nm light illumination, the diodes showed a responsivity close to 104 A W−1. These initial results demonstrate the potential of the SSNDs toward monolithic Ga2O3 circuits that are currently under extensive investigation for various applications....
This work presents the performance projection of a metal-insulator-graphene diode as the building block of a radiofrequency power detector, highlighting its rectifying figures of merit. The analysis was performed by means of a computer-aided design tool validated with experimental measurements of fabricated devices. Transient simulations were used to accurately determine the detector output voltage, while particular consideration was given to suitable convergence of the non-linear circuit response. The diode was analyzed in both ideal and non-ideal cases, with the latter accounting for its parasitic effects. In the non-ideal case, the diode exhibited a tangential responsivity of 26.9 V/W at 2.45 GHz and 31.9 V/W at 1.225 GHz. However, when parasitic elements were neglected in the ideal case, the responsivity significantly increased to 47.3 V/W at 2.45 GHz and 38.7 V/W at 1.225 GHz. Additionally, the diode demonstrated a non-linearity of 6.64 at 0.7 V and an asymmetry of 806.6 in a bias window of ±1 V, which resulted in a competitive value compared to other state-of-the-art rectifying technologies. Tangential responsivities (βv) of graphene diodes at less-studied frequencies in the gigahertz band are presented, showing a high βv value of 63.7 V/W at 1 GHz....
Loading....