While gaining significant attention for device-free applications, Wi-Fi sensing still faces challenges in differentiating multiple targets; this stems from the design priorities of Wi-Fi systems that prioritize coverage and stability over sensing diversity. Existing proposals that either expand bandwidth or increase antennas to enhance sensing diversity can be confined by the limited access to Wi-Fi firm/hardware. To this end, we propose Mirror-Fi, a novel Wi-Fi sensing system that improves sensing diversity without modifying Wi-Fi firm/hardware. Exploiting the reconfigurability of metasurfaces, Mirror-Fi augments beamforming with spatially significant features, facilitating the construction of exclusive sensing signal links for individual targets. We innovate in an encoding scheme that equips each metasurface with a distinct phase coding sequence to mark link uniqueness. We then train a deep neural model to leverage prior coding sequences for decomposing non-linearly superimposed channel samples into mutually independent channels; it removes the need for complex channel matrix parameter estimation and mitigates hardware-related offsets inherent to Wi-Fi. Extensive evaluations demonstrate that, with a sufficient number of auto-configured metasurfaces, Mirror-Fi successfully achieves multi-target sensing.

Millimeter-wave (mmWave) radar offers significant potential for fine-grained sensing, yet transitioning from controlled laboratory environments to dynamic real-world applications remains challenging. Existing methods face a dichotomy: real-time point clouds sacrifice crucial signal details needed for sophisticated tasks, whereas information-rich raw data sensing imposes prohibitive transmission and computation overheads, often limiting analysis to offline settings and hindering real-time viability. To this end, we present CoSense, areal-timeedge-end collaborative sensing system built on commodity mmWave radar (end) and edge intelligence. We first introduce a novel dual-stream data acquisition mechanism via realizing radar driver-level interfaces, enabling simultaneous transmission of point clouds and raw data. To bridge the fidelity-latency trade-off, we implement an adaptive transmission strategy via firmware modifications, selectively forwarding raw data segments (corresponding to regions of interest identified in the point cloud) for detailed fine-grained analysis, while continuously delivering point clouds for low-latency coarse-grained sensing and control loops. Furthermore, we incorporate closed-loop feedback beamforming, dynamically steering the radar beam based on real-time tracking to counteract motion-induced misalignment and enhance signal fidelity. Extensive evaluations under dynamic conditions demonstrate that CoSense successfully achieves real-time fine-grained sensing with high fidelity and manageable overhead.

Speech enhancement is crucial for reliable communication in noisy environments. However, the lack of a priori knowledge about target speech characteristics in conventional systems often leads to erroneous extraction of interfering speech as desired signals during noise suppression, significantly compromising system performance. Recently, researchers have proposed to use the side-channel signal as an assist to denoise noisy speech. This paper proposes BoneSE, a multimodal speech enhancement approach using bone conduction. The basic idea of BoneSE is to perceive the bone-conducted sound with an IMU sensor embedded in the Commercial Off-The-Shelf (COTS) earphone and then leverage the correlations between the bone conduction signal and audio signal for speech enhancement. However, the lack of high-frequency components in the IMU modality brings data imbalance and hinders data fusion. To address this challenge, we explore and model the relationship between multimodal signals and design a fusion module according to the time and frequency correlation. Moreover, to balance fast processing and denoising performance, we propose bone conduction-based noise level metrics to measure the noise level. To accommodate different noise levels, we propose an adaptive model selection approach based on reinforcement learning to select the proper denoising model, thereby optimizing the latency. Experiments on two datasets show that the proposed method performs favorably against state-of-the-art methods and can enhance speech in low SNR scenarios.

Human pose recognition is a crucial task in millimeter-wave point cloud processing, playing an important role in scenarios such as autonomous driving, human-computer interaction, and medical monitoring. Currently, most approaches for recognizing human poses from millimeter-wave point clouds rely on deep learning methods. However, due to the impact of multipath effects in wireless signals, millimeter-wave point clouds not only contain human pose information but also include environmental noise. The neural network extracts features from all point clouds equally, which leads to errors in human pose recognition caused by environmental noise interference. To address this limitation, we propose TSDA, a novel Temporal-Spatial Data Augmentation method specifically tailored for human pose recognition in millimeter-wave point clouds. TSDA leverages the temporal consistency and spatial continuity of human pose point clouds to distinguish the real human point clouds from environmental noise point clouds. It quantifies the probability of a point cloud belonging to the human pose target through confidence measures, which can be used to augment the raw point clouds. Additionally, we introduce a prototype system based on cross-attention mechanisms to validate the impact of data augmentation on pose recognition performance. Experimental results show that, compared to the state-of-the-art deep learning-based methods, TSDA achieves an average reduction of 1.9cm in human pose recognition error.

Research on protein stability changes is vital for understanding disease mechanisms and optimizing industrial enzymes. Protein thermal stability can be modified by variants leading to changes in ΔΔG values between wild-type and mutant proteins. Despite advances, most models focus on single-point mutations, overlooking multipoint and indel mutations. Typically, the single-point mutation is expected to have a relatively limited impact on the function of a protein, necessitating more drastic modifications to meet new challenges. Current methods for multipoint mutations yield poor results, and no method exists for any length of indel mutations. To address this, we introduce UniMutStab, a shared-graph convolutional network leveraging protein language models and residue interaction networks to access any type of mutation. An embedded edge weight module enhances the integration of residue node features and interactions, improving prediction accuracy. Trained on the "Mega-scale" dataset with ~780 000 mutations, UniMutStab surpasses existing methods in predicting protein stability changes. It is a purely sequence-based approach to predict arbitrary mutation types, demonstrating robust generalization across multiple tasks and potentially contributing significantly to protein engineering, personalized therapeutics, and diagnostic methodologies.

The humanization of antibodies (Abs) remains one of the main pathways for therapeutic antibody development. With the advantages of diffusion models, here we present HuAbDiffusion, a discrete language diffusion model used for antibody humanization by generating humanized antibodies from scratch. HuAbDiffusion starts from three complementary determinant regions (CDRs) and finally generate whole V region sequences. The model was evaluated on 22 mAbs and compared with several existing methods, the test results show the effectiveness and better performance of the proposed model. Besides, the potential optimal humanized antibodies to be selected could be narrowed down to a reasonable level with the usage of pretrained language models. The most significant is that the binding affinity of the humanized antibody can be retained or even increased generated by HuAbDiffusion. The method can be reached out through our previous established YabXnization server at https://www.genscript.com/tools/yabxnization-service.

With the rise of intelligent systems like assisted driving and robotics, all-weather target identification and 3D localization systems have become crucial for reliable obstacle avoidance and navigation. However, vision-based methods struggle to provide accurate target locations under low light or bad weather. Radar-based solutions like mmWave radar and LiDAR are robust but hindered by high costs and challenges in recognizing target identities at scale. In this paper, we propose alow-cost, all-weather target identification and 3D localization systembased onLED-tags, which system can address the needs of intelligent systems for obstacle avoidance in complex environments. We explore the backscatter communication of LED devices and design adual-modal LED-Tag, which includes two features: a backscatter RF signal detectable by RF devices and visual light spot information detectable by cameras, both sharing the same ID. To enhance the limited backscatter capability, we propose amulti-branch parallel modelthat enhances the signal strength using beamforming synthesis and achannel adjustment mechanismto improve robustness in complex environments, ensuring accurate 3D localization. For multi-target identification, we design an LED-tag encoding system, assigning each tag a unique encoding sequence. Each target's identity can be recognized with our customizedID decoding method, which leverages prior information and time-domain sampling characteristics. Extensive experimental results show that the backscatter communication and target detection range of LED-tags can reach15m. Moreover, the system achieves anaverage localization error of 7.3cm within a 5m range, demonstrating the system's excellent performance in terms of practicality and accuracy.

Nowadays, object detection and 3D tracking have become key technologies for intelligent system or robot navigation to realize automatic obstacle avoidance and target detection, especially in low-light and night vision scenarios. In this paper, we explore the backscattering capability of LEDs and implement a multi-modal tag LED-tag to realize object detection and 3D tracking. Our basic idea is to utilize the fact that feeding modulation signals to an LED-tag can generate both RF and visual features. We fuse the depth of field information perceived from the RF domain and the pixel coordinates obtained from the visual domain to derive a 3D position by matching the decoded ID. In the RF domain, the depth of field is acquired through ultra-wideband channel measurements and estimated phase. In the visual domain, the pixel coordinate in the XOY coordinates can be extract from the image and mapped into 2D spatial coordinates. To address the limited backscatter capability of the LED-tag, we propose a multiple parallel branch model to increase backscatter paths for amplifying the LED-tag's backscattering intensity. Additionally, we propose a decoding ID scheme that utilizes a priori knowledge and repetitive samples to restore the IDs whose encoding frequency is higher than four times the sampling rate. We have implemented a prototype system and evaluated its performance in real-world environments. Extensive experimental results show that LED-tag can backscatter RF signals ranging up to 15m. Besides, the system achieves an average position error of 8cm within the range of 3m.

BACKGROUND: The study of codon usage bias is important for understanding gene expression, evolution and gene design, providing critical insights into the molecular processes that govern the function and regulation of genes. Codon Usage Bias (CUB) indices are valuable metrics for understanding codon usage patterns across different organisms without extensive experiments. Considering that there is no one-fits-all index for all species, a comprehensive platform supporting the calculation and analysis of multiple CUB indices for codon optimization is greatly needed. RESULTS: Here, we release GenRCA, an updated version of our previous Rare Codon Analysis Tool, as a free and user-friendly website for all-inclusive evaluation of codon usage preferences of coding sequences. In this study, we manually reviewed and implemented up to 31 codon preference indices, with 65 expression host organisms covered and batch processing of multiple gene sequences supported, aiming to improve the user experience and provide more comprehensive and efficient analysis. CONCLUSIONS: Our website fills a gap in the availability of comprehensive tools for species-specific CUB calculations, enabling researchers to thoroughly assess the protein expression level based on a comprehensive list of 31 indices and further guide the codon optimization.

With the proliferation of voice assistants, microphone-based speech recognition technology usually cannot achieve good performance in the situation of multiple sound sources and ambient noises. In this paper, we propose a novel mmWave-based solution to perform speech recognition to tackle the issues of multiple sound sources and ambient noises, by precisely extracting the multi-modal features from lip motion and vocal-cords vibration from the single channel of mmWave. We propose a difference-based method for feature extraction of lip motion to suppress the dynamic interference from body motion and head motion. We propose a speech detection method based on cross-validation of lip motion and vocal-cords vibration so as to avoid wasting computing resources on nonspeaking activities. We propose a multi-modal fusion framework for speech recognition by fusing the signal features from lip motion and vocal-cords vibration with the attention mechanism. We implemented a prototype system and evaluated the performance in real test-beds. Experiment results show that the average speech recognition accuracy is 92.8% in realistic environments.

This paper mainly simulated the single event latch-up (SEL) for the CMOS inverter under the 0.18um technology. The SEL of integrated circuit (IC) was also analyzed in detail. The result showed that the parasitic lateral transistors NPN and PNP of NMOS and PMOS play a role in the SEL happening process. The changes of the drain voltage and the drain current and the functional failure of the circuit were also explained in further. Therefore the new SEL model could be established.

The paper simulated the SEL happening process of the CMOS inverter fabricated the 0.18um technology. The results show that the intrinsic parasitic lateral NPN (QN) and PNP (QP) transistor of the NMOS and PMOS in the CMOS inverter, which could result in the changes of the voltage and the current of the drain when the SEL happening, can delay latch up occurring time and reduce the latch up current. The origin model was improved based on the simulated results. The result studying the improved latch up model shows that the smaller ratios of the internal parasitic resistors between RW1and RW2or RS1and RS2could lead to smaller latch up current and delay more time for the latch up occurrence.

This paper presents an improved tool called FITVS (Fault Injection Tool for Validating SEE) using the FPGA-based emulation system for fault grading. A novel library-replace-modeling technique that can quickly and easily perform SEE by injecting faults into the circuit nodes is proposed. It helps IC designers to enhance the quality of their design by providing the sensitivity information of all nodes. Also the fault injection effectiveness is improved with relative to the traditional methods by utilizing C# program and FPGA emulation, and the speed of injection can reach the order of lus/fault.