Gate Modulation Research Articles

Advancing Rural Building Extraction via Diverse Dataset Construction and Model Innovation with Attention and Context Learning

Rural building automatic extraction technology is of great significance for rural planning and disaster assessment; however, existing methods face the dilemma of scarce sample data and large regional differences in rural buildings. To solve this problem, this study constructed an image dataset of typical Chinese rural buildings, including nine typical geographical regions, such as the Northeast and North China Plains. Additionally, an improved remote sensing image rural building extraction network called AGSC-Net was designed. Based on an encoder–decoder structure, the model integrates multiple attention gate (AG) modules and a context collaboration network (CC-Net). The AG modules realize focused expression of building-related features through feature selection. The CC-Net module models the global dependency between different building instances, providing complementary localization and scale information to the decoder. By embedding AG and CC-Net modules between the encoder and decoder, the model can capture multiscale semantic information on building features. Experiments show that, compared with other models, AGSC-Net achieved the best quantitative metrics on two rural building datasets, verifying the accuracy of the extraction results. This study provides an effective example for automatic extraction in complex rural scenes and lays the foundation for related monitoring and planning applications.

GCRec: Graph-Augmented Capsule Network for Next-Item Recommendation.

Next-item recommendation has been a hot research, which aims at predicting the next action by modeling users' behavior sequences. While previous efforts toward this task have been made in capturing complex item transition patterns, we argue that they still suffer from three limitations: 1) they have difficulty in explicitly capturing the impact of inherent order of item transition patterns; 2) only a simple and crude embedding is insufficient to yield satisfactory long-term users' representations from limited training sequences; and 3) they are incapable of dynamically integrating long-term and short-term user interest modeling. In this work, we propose a novel solution named graph-augmented capsule network (GCRec), which exploits sequential user behaviors in a more fine-grained manner. Specifically, we employ a linear graph convolution module to learn informative long-term representations of users. Furthermore, we devise a user-specific capsule module and a position-aware gating module, which are sensitive to the relative sequential order of the recently interacted items, to capture sequential patterns at union-level and point-level. To aggregate the long-term and short-term user interests as a representative vector, we design a dual-gating mechanism, which could decide the contribution ratio of each module given different contextual information. Through extensive experiments on four benchmarks, we validate the rationality and effectiveness of GCRec on the next-item recommendation task.

Breaking the responsivity-speed dilemma of a-GaOx photodetector by alternating gate modulation

Breaking the responsivity-speed dilemma of a-GaOx photodetector by alternating gate modulation

Evaluation of dual-ONOFIC method for subthreshold leakage reduction in domino circuit

ABSTRACT A novel dual ON/OFF logic (ONOFIC) pull-down method is presented to reduce subthreshold leakage current in FinFET domino circuit wide OR gates at the 32 nm technology node. An ONOFIC block is inserted in the pull-down network of dynamic and inverted blocks. The HSPICE simulator with the 32 nm BISM4 model is used for the simulation of the proposed work. The outcome of the CLIL (clock is low and inputs are low) state is most effective in reducing the subthreshold leakage current at low and high temperatures. The OR2, OR4, OR8, and OR16 circuits using the proposed method cut down the subthreshold leakage power dissipation up to 48.1% for low power (LP) mode when compared to the LECTOR domino gates in CHIL (clock is high and inputs are low) state, which also reduced subthreshold leakage power dissipation up to 74% for short gate (SG) mode. Due to a lower subthreshold current when all inputs are low at low and high temperatures, the proposed dual-ONOFIC pull-down technique outperforms the ONOFIC pull-up technique in LP mode. When inputs are high at low and high temperatures, the ONOFIC pull-up technique performs better than the dual-ONOFIC pull-down technique.

Exploiting nonlinear dendritic adaptive computation in training deep Spiking Neural Networks

Inspired by the information transmission process in the brain, Spiking Neural Networks (SNNs) have gained considerable attention due to their event-driven nature. However, as the network structure grows complex, managing the spiking behavior within the network becomes challenging. Networks with excessively dense or sparse spikes fail to transmit sufficient information, inhibiting SNNs from exhibiting superior performance. Current SNNs linearly sum presynaptic information in postsynaptic neurons, overlooking the adaptive adjustment effect of dendrites on information processing. In this study, we introduce the Dendritic Spatial Gating Module (DSGM), which scales and translates the input, reducing the loss incurred when transforming the continuous membrane potential into discrete spikes. Simultaneously, by implementing the Dendritic Temporal Adjust Module (DTAM), dendrites assign different importance to inputs of different time steps, facilitating the establishment of the temporal dependency of spiking neurons and effectively integrating multi-step time information. The fusion of these two modules results in a more balanced spike representation within the network, significantly enhancing the neural network’s performance. This approach has achieved state-of-the-art performance on static image datasets, including CIFAR10 and CIFAR100, as well as event datasets like DVS-CIFAR10, DVS-Gesture, and N-Caltech101. It also demonstrates competitive performance compared to the current state-of-the-art on the ImageNet dataset.

The dominant-negative effect of Loss of function Nav 1.5 variants is not a critical determinant of phenotype severity

Abstract Introduction Loss-of-function (LOF) variants of SCN5A cause various arrhythmic diseases. Recent studies have demonstrated that some LOF-SCN5A variants can exert a dominant-negative (DN) effect through channel dimerization, resulting in higher penetrance of disease phenotypes. However, the association between the DN effect and phenotype severity remains unclear. In this study, we identified a proband with compound heterozygous SCN5A variants, p.G833R and p.T1396P, in a family with bradyarrhythmia. We analyzed the electrophysiological function of these variants and their association with channel-dimerization to elucidate their effects on phenotype severity. Purpose The purpose of this study is to investigate the effect of sodium channel dimerization on the phenotype severity in a family with bradyarrhythmia Methods We performed functional analysis of the mutant Nav1.5s induced in HEK cells utilizing a whole-cell patch-clamp method. The DN effect was also assessed by co-expressing WT and mutants in each heterozygous combinations, with or without difopein, a peptide working as a channel-coupling canceler. Cell surface expression of mutated channels were measured by flowcytometry. Results We identified SCN5A p.T1396P and p.G833R variants segregating in a family suffering from severe bradyarrhythmia. The proband carried both variants in a compound heterozygous manner, while her mother carried only p.T1396P. The clinical phenotype was more severe and had early onset in the proband; the compound heterozygous case. In the patch-clamp study, HEK 293 cells solely expressing T1396P did not show any measurable sodium current, while those with G833R had comparative current with that of WT channel. Cells expressing WT/T1396P showed significant reduction and delay of inactivation decay, which was normalized by co-expression with difopein. The results implied that loss-of-function variant T1396P manifested DN-effect via channel dimerization. Cells co-expressing T1396P and G833R did not demonstrate a DN-effect, suggesting that Nav1.5-G833R did not form a dimer. Considering more severe phenotype of compound heterozygous member in the family, the DN-cancelation by monomerized Nav1.5-G833R was not associated with a mild phenotype. Conclusion Our result supports the current Nav1.5-dimerization concept and consequent DN-effect by LOF-SCN5A variants. However, the phenotype severity was not simply explainable by the presence of a dominant negative effect. Further multiplanar analyses including assays for channel clustering, localization, and gating modification would be necessary.

Wikstroemia indica (L.) C. A. Mey. Exerts analgesic activity by inhibiting NaV1.7 channel

Ethnopharmacological relevanceWikstroemia indica (L.) C. A. Mey. is traditionally used for the treatment of gastrointestinal disorders, respiratory illnesses, skin infections, and inflammatory conditions. Despite extensive evidence of its biological potential, including antipyretic, antimicrobial, antifungal, anti-inflammatory, and diuretic properties, there are currently no reports indicating its analgesic effects. Aim of the studyCrude extracts from W. indica stems were examined for anti-nociceptive activity. Additionally, an in-depth investigation was conducted to uncover the molecular basis for the possible analgesic phenomenon. Materials and methodsW. indica stems were subjected to ethanol extraction. To evaluate the in vivo analgesic activity, both chemical and physical-induced pain models were employed. Additionally, single-cell electrophysiological recordings were performed on human embryonic kidney 293T (HEK293T) cells expressing NaV1.7 channel. ResultsCrude extracts derived from W. indica exhibited significant efficacy in mitigating the pain sensation, as evidenced by their substantial effects in both acetic acid-induced and heat-induced pain models. Further screening unveiled osthenol as a key bioactive compound responsible for mediating the analgesic properties of W. indica. Osthenol directly interacts with the pore domain of NaV1.7 channels, leading to channel inhibition. Importantly, this interaction is independent of any changes in the channel gating modifier domain. ConclusionBoth W. indica and osthenol demonstrate potential as effective anti-nociceptive agents in preclinical studies. Their analgesic effects are likely achieved by inhibiting the NaV1.7 channel, which is crucial in pain initiation, transmission, and modulation. These results elucidate the molecular basis of the W. indica as a pain-relieving medication. Additionally, osthenol holds great potential in advancing the development of anti-nociceptive drugs targeting the NaV1.7 channel.

A lightweight multi-modality medical image semantic segmentation network base on the novel UNeXt and Wave-MLP

Medical images sometimes contain diseased regions that are different sizes and. shapes, which makes it difficult to accurately segment these areas or their edges. However, directly coupling CNN and MLP to construct global and local dependency. models may also cause significant computational complexity issues. In this paper, a. unique, lightweight UNeXt network segmentation model for medical images based on. dynamic aggregation tokens was proposed. Firstly, the Wave Block module in Wave-MLP was introduced to replace the Tok-MLP module in UNeXt. The phase term in Wave Block can dynamically aggregate tokens, improving the segmentation accuracy of the model. Secondly, an AG attention gate module is added at the skip connection to suppress irrelevant feature representations in the sampling path of the encoding. network, thereby reducing computational costs and paying attention to noise and artifacts. Finally, the Focal Tversky Loss was added to handle both binary and multiple classification jobs. Quantitative and qualitative experiments were conducted on two public datasets: COVID-19 CT and BraTS 2018 MRI. The Dice score, Precision score, recall score, and Iou score of the proposed model on the COVID-19 dataset were 0.928, 0.867, 0.916, and 0.940, respectively. On BraTS 2018, the Dice scores of the ET, WT, and TC categories were 0.933, 0.925, and 0.918, respectively, and the HD scores were 1.595, 2.348, and 1.549, respectively. At the same time, the model is lightweight and has a considerably decreased training time with GFLOPs and Params of 0.52 and 0.76, respectively. The proposed lightweight model is superior to other existing methods in terms of segmentation accuracy and computing complexity according to experimental data.

Multitype Perception Method for Drug-Target Interaction Prediction.

With the growing popularity of artificial intelligence in drug discovery, many deep-learning technologies have been used to automatically predict unknown drug-target interactions (DTIs). A unique challenge in using these technologies to predict DTI is fully exploiting the knowledge diversity across different interaction types, such as drug-drug, drug-target, drug-enzyme, drug-path, and drug-structure types. Unfortunately, existing methods tend to learn the specifical knowledge on each interaction type and they usually ignore the knowledge diversity across different interaction types. Therefore, we propose a multitype perception method (MPM) for DTI prediction by exploiting knowledge diversity across different link types. The method consists of two main components: a type perceptor and a multitype predictor. The type perceptor learns distinguished edge representations by retaining the specifical features across different interaction types; this maximizes the prediction performance for each interaction type. The multitype predictor calculates the type similarity between the type perceptor and predicted interactions, and the domain gate module is reconstructed to assign an adaptive weight to each type perceptor. Extensive experiments demonstrate that our proposed MPM outperforms the state-of-the-art methods in DTI prediction.

Predicting the miniaturization limit of vertical organic field effect transistor (VOFET) with perforated graphene as a source electrode

Vertical organic field effect transistors (VOFETs) are of paramount importance due to their fast switching speed, low power consumption, and higher density on a chip compared to lateral OFETs. The low charge carrier mobility in organic semiconductors and longer channel lengths in lateral OFETs lead to higher operating voltages. The channel length in VOFETs can be less than 100 nm which reduces the size of the channel and hence the operating voltages. Another important factor in the operation of VOFETs is the thickness and width of the source electrode. The channel length, source electrode thickness and width sets the miniaturization limit of the VOFETs. The graphene monolayer can be exploited as a source electrode due to its thinness, high carrier mobility, and metallic behaviors. However, for better gate modulation, perforations in the source material are desired. Here, we simulate the VOFET having perforated graphene monolayer as a source electrode and n-type organic semiconductor N, N′-dioctyl-3,4,9,10-perylenedicarboximide (PTCDI-C8) as an active channel material, while aluminum as a drain electrode to predict the best-miniaturized device. The miniaturization limit of such a VOFET has a limit to the gate opening/perforation in which the minimum source width is 10 nm, as in the sub 10 nm range graphene starts behaving like a semiconductor. The subthreshold swing, deduced from the drain current (J D) versus gate voltage (V G) graph, advocates the limit of the organic semiconductor height/channel length to 50 nm, while 50 nm for the gate.

SUnet: A multi-organ segmentation network based on multiple attention

Organ segmentation in abdominal or thoracic computed tomography (CT) images plays a crucial role in medical diagnosis as it enables doctors to locate and evaluate organ abnormalities quickly, thereby guiding surgical planning, and aiding treatment decision-making. This paper proposes a novel and efficient medical image segmentation method called SUnet for multi-organ segmentation in the abdomen and thorax. SUnet is a fully attention-based neural network. Firstly, an efficient spatial reduction attention (ESRA) module is introduced not only to extract image features better, but also to reduce overall model parameters, and to alleviate overfitting. Secondly, SUnet's multiple attention-based feature fusion module enables effective cross-scale feature integration. Additionally, an enhanced attention gate (EAG) module is considered by using grouped convolution and residual connections, providing richer semantic features. We evaluate the performance of the proposed model on synapse multiple organ segmentation dataset and automated cardiac diagnostic challenge dataset. SUnet achieves an average Dice of 84.29% and 92.25% on these two datasets, respectively, outperforming other models of similar complexity and size, and achieving state-of-the-art results.

Alterations of auditory sensory gating in mice with noise-induced tinnitus treated with nicotine and cannabis extract.

Tinnitus is a phantom sound perception affecting both auditory and limbic structures. The mechanisms of tinnitus remain unclear and it is debatable whether tinnitus alters attention to sound and the ability to inhibit repetitive sounds, a phenomenon also known as auditory gating. Here we investigate if noise exposure interferes with auditory gating and whether natural extracts of cannabis or nicotine could improve auditory pre-attentional processing in noise-exposed mice. We used 22 male C57BL/6J mice divided into noise-exposed (exposed to a 9-11 kHz narrow band noise for 1 h) and sham (no sound during noise exposure) groups. Hearing thresholds were measured using auditory brainstem responses, and tinnitus-like behavior was assessed using Gap prepulse inhibition of acoustic startle. After noise exposure, mice were implanted with multi-electrodes in the dorsal hippocampus to assess auditory event-related potentials in response to paired clicks. The results showed that mice with tinnitus-like behavior displayed auditory gating of repetitive clicks, but with larger amplitudes and longer latencies of the N40 component of the aERP waveform. The combination of cannabis extract and nicotine improved the auditory gating ratio in noise-exposed mice without permanent hearing threshold shifts. Lastly, the longer latency of the N40 component appears due to an increased sensitivity to cannabis extract in noise-exposed mice compared to sham mice. The study suggests that the altered central plasticity in tinnitus is more sensitive to the combined actions on the cholinergic and the endocannabinoid systems. Overall, the findings contribute to a better understanding of pharmacological modulation of auditory sensory gating.

A high-throughput electrophysiology assay to study the response of PIEZO1 to mechanical stimulation.

PIEZO1 channels are mechanically activated cation channels that play a pivotal role in sensing mechanical forces in various cell types. Their dysfunction has been associated with numerous pathophysiological states, including generalized lymphatic dysplasia, varicose vein disease, and hereditary xerocytosis. Given their physiological relevance, investigating PIEZO1 is crucial for the pharmaceutical industry, which requires scalable techniques to allow for drug discovery. In this regard, several studies have used high-throughput automated patch clamp (APC) combined with Yoda1, a specific gating modifier of PIEZO1 channels, to explore the function and properties of PIEZO1 in heterologous expression systems, as well as in primary cells. However, a combination of solely mechanical stimulation (M-Stim) and high-throughput APC has not yet been available for the study of PIEZO1 channels. Here, we show that optimization of pipetting parameters of the SyncroPatch 384 coupled with multihole NPC-384 chips enables M-Stim of PIEZO1 channels in high-throughput electrophysiology. We used this approach to explore differences between the response of mouse and human PIEZO1 channels to mechanical and/or chemical stimuli. Our results suggest that applying solutions on top of the cells at elevated pipetting flows is crucial for activating PIEZO1 channels by M-Stim on the SyncroPatch 384. The possibility of comparing and combining mechanical and chemical stimulation in a high-throughput patch clamp assay facilitates investigations on PIEZO1 channels and thereby provides an important experimental tool for drug development.

Multi-scale Bottleneck Residual Network for Retinal Vessel Segmentation.

Precise segmentation of retinal vessels is crucial for the prevention and diagnosis of ophthalmic diseases. In recent years, deep learning has shown outstanding performance in retinal vessel segmentation. Many scholars are dedicated to studying retinal vessel segmentation methods based on color fundus images, but the amount of research works on Scanning Laser Ophthalmoscopy (SLO) images is very scarce. In addition, existing SLO image segmentation methods still have difficulty in balancing accuracy and model parameters. This paper proposes a SLO image segmentation model based on lightweight U-Net architecture called MBRNet, which solves the problems in the current research through Multi-scale Bottleneck Residual (MBR) module and attention mechanism. Concretely speaking, the MBR module expands the receptive field of the model at a relatively low computational cost and retains more detailed information. Attention Gate (AG) module alleviates the disturbance of noise so that the network can concentrate on vascular characteristics. Experimental results on two public SLO datasets demonstrate that by comparison to existing methods, the MBRNet has better segmentation performance with relatively few parameters.

A Power-Gated 8-Transistor Physically Unclonable Function Accelerates Evaluation Speeds

The proposed 8-Transistor (8T) Physically Unclonable Function (PUF), in conjunction with the power gating technique, can significantly accelerate a single evaluation cycle more than 100,000 times faster than a 6-Transistor (6T) Static Random-Access Memory (SRAM) PUF. The 8T PUF is built to swiftly eliminate data remanence and maximise physical mismatch. Moreover, a two-phase power gating module is devised to provide controllable power on/off cycles for the chosen PUF clusters in order to facilitate fast statistical measurements and curb the in-rush current. The architecture and hardware implementation of the power-gated PUF are developed to accommodate fast multiple evaluations of PUF Responses. The fast speed enables a new data processing method, which coordinates Dark-bit masking and Multiple Temporal Majority Voting (TMV) in different Process, Voltage and Temperature (PVT) corners or during field usage, hence greatly reducing the Bit Error Rate (BER) and the hardware penalty for error correction. The designs are based on the UMC 65 nm technology and aim to tape out an Application-Specific Integrated Circuit (ASIC) chip. Post-layout Monte Carlo (MC) simulations are performed with Cadence, and the extracted PUF Responses are processed with Matlab to evaluate the 8T PUF performance and statistical metrics for subsequent inclusion in PUF Responses, which comprise the novelty of this approach.

Enhancing Clinical Accuracy in DIBH Pattern Detection with a Combined Ultrasonic and IR Sensor System

This study aimed to develop a DIBH monitoring system using IR (Infrared- VL6180X Time of Flight) distance sensors and a microcontroller, and to demonstrate its functionality and performance in various irregular free and DIBH breathings from the left breast and lung patients. A total of ten patients who performed DIBH breathing for treatment planning and optimized treatment delivery were randomly selected for the study. The breathing motion of the system was tested using an RPM (Real-Time Positioning Management) phantom. The IR sensors were interfaced with an Arduino microcontroller, and a body phantom was used to mimic the DIBH pattern on a CT scan. The detector holder was mounted perpendicularly between the xiphoid and umbilicus, with the ability to compensate for various patient body habitus and adapt to clinical setups. The plastic devices were 3D printed, and the sensors were mounted approximately 100mm apart to receive directed echo signals and prevent unwanted signals from scattering. The IR sensor DIBH system was able to detect and record the patient's breathing pattern in real-time. The IR sensor showed better visualization of breathing rhythm with fewer fluctuations and reduced noise than the previous ultrasonic sensor system. The breathing amplitude and duration for the patients were compared to our treatment planning of Varian Eclipse and verified in TrueBeam's Breath-Hold Gating mode. The amplitude and duration from the primary study were measured to ±5mm, including 3mm sensor noise level and ±1.5 second. The IR sensor DIBH system demonstrated better performance compared to the ultrasonic sensor, with a better visualization of breathing rhythm and reduced fluctuations and noise. It provided a clinically acceptable DIBH pattern for monitoring chest and abdominal motion in patients with irregular breathing. The findings of this study have the potential to enhance clinical accuracy in radiation therapy treatment planning and delivery for patients undergoing DIBH.

HCMS: Hierarchical and Conditional Modality Selection for Efficient Video Recognition

Videos are multimodal in nature. Conventional video recognition pipelines typically fuse multimodal features for improved performance. However, this is not only computationally expensive but also neglects the fact that different videos rely on different modalities for predictions. This article introduces Hierarchical and Conditional Modality Selection (HCMS), a simple yet efficient multimodal learning framework for efficient video recognition. HCMS operates on a low-cost modality, i.e., audio clues, by default, and dynamically decides on-the-fly whether to use computationally expensive modalities, including appearance and motion clues, on a per-input basis. This is achieved by the collaboration of three LSTMs that are organized in a hierarchical manner. In particular, LSTMs that operate on high-cost modalities contain a gating module, which takes as inputs lower-level features and historical information to adaptively determine whether to activate its corresponding modality; otherwise, it simply reuses historical information. We conduct extensive experiments on two large-scale video benchmarks, FCVID and ActivityNet, and the results demonstrate the proposed approach can effectively explore multimodal information for improved classification performance while requiring much less computation.

MIFINN: A novel multi-information fusion and interaction neural network for aspect-based sentiment analysis

Aspect-based Sentiment Analysis (ABSA) is used to detect corresponding sentiment polarities toward different aspect terms. In recent years, graph convolutional network-based methods achieve great success for ABSA. However, previous studies ignore syntactic dependency types and aspect-related syntactic distances. Besides, effectively fusing semantic and syntactic information remains a challenging problem. Previous studies also neglect interactions between aspect terms and context. To alleviate the above problems, we develop a Multi-Information Fusion and Interaction Neural Network (MIFINN), which consists of an Aspect and Syntax-Aware Multi-Information Fusion Graph Convolutional Network (ASAMIFGCN) and a Multi-Information Interaction and Gating Network (MIIGN). The ASAMIFGCN model can integrate and fuse contextual semantics, syntactic dependency types, and aspect-related syntactic distances. Specifically, two different attentions are designed to learn global and aspect-related semantic and syntactic information, respectively. A syntactic distance gating module is proposed to acquire more precise score matrices. Besides, we propose a semantic and syntactic association module to fuse multi-information simultaneously. The MIIGN model can effectively interact aspect terms and context. Specifically, two interactive attentions are developed to learn different weights toward aspect terms and context via interactions between each other, respectively. Moreover, we propose a gating information module for controlling the information flow of aspect terms and context. We conduct experiments on multiple datasets and MIFINN achieves state-of-the-art performance. The results also demonstrate that effectively fusing and interacting multi-information can improve model performance for ABSA significantly.

Response Fusion Attention U-ConvNext for accurate segmentation of optic disc and optic cup

Glaucoma is a pathological eye condition which requires accurate optic disc and optic cup segmentation for diagnosis. This study proposes Responsive Fusion Attention U-ConvNext, a novel encoder decoder network, for semantic segmentation of optic disc and optic cup structures from fundus images. Response Fusion Attention U-ConvNext is a U-Net like model containing a pre-trained ConvNext encoder network and a light-weight, modified ConvNext decoder network having skip connections between the encoder and decoder blocks. We also propose a new attention gate module called Dual-Path Response Fusion Attention (DPRFA) for smoothing the concatenation process of the encoder and upsampled feature maps. In addition, we propose a modified loss function by combining cross entropy, Dice and Jaccard losses for training the model accurately. The model has four sizes and all of them are trained and validated on DRISHTI-GS and REFUGE datasets. Our proposed models have acheived a dice coefficient of 0.9822 and 0.9269 on optic disc and optic cup segmentation of DRISHTI-GS dataset and a dice coefficient of 0.9788 and 0.9086 on optic disc and optic cup segmentation of REFUGE dataset. The experimental results thus obtained by our suggested models have shown state-of-the-art results when compared to other existing models. Code and models are available at : https://github.com/SiddMallick/RFAUCNxt-official.

An Ultrasensitive ReSe2/WSe2 Heterojunction Photodetector Enabled by Gate Modulation and its Development in Polarization State Identification

AbstractUltrasensitive photodetectors with polarization angle recognition have broad applications in both civilian and military domains. The emerging 2D materials with in‐plane anisotropy offer promising platforms for realizing these applications, owing to their intriguing properties. However, the lack of an effective photoconductivity gain mechanism and the low photocurrent anisotropic ratio have made it challenging to achieve digital output of polarization information. Herein, a gate voltage‐dependent phototransistor based on ReSe2/WSe2 out‐of‐plane heterostructure is proposed. Attributed to sophisticated band alignment engineering, the device exhibits remarkable photoresponse characteristics and polarization sensitivity, including an outstanding responsivity of 2.29 × 104 A W−1, an extraordinary detectivity of 5.65 × 1013 Jones, a prompt rise/decay time of 197/182 µs, as well as a high photocurrent anisotropic ratio of 10.9. Based on these features, the device is integrated into a polarization recognition system, which accurately identifies the polarization angle of the incident light and displays it. These results demonstrate the strong potential of the ReSe2/WSe2 device for future applications in polarization optical communication and logic circuits.