Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks251 from public data
- The Visual Object Tracking VOT2016 Challenge Results776
The Visual Object Tracking challenge VOT2016 goes beyond its predecessors by introducing a new semi-automatic ground truth bounding box annotation methodology and extending the evaluation system with the no-reset experiment.
- 678
- Ternary Weight Networks604
Extensive experiments show that the TWNs achieve much better result than the Binary-Weight-Networks (BWNs) and the classification performance on MNIST and CIFAR-10 is very close to the full precision networks.
- Cross-Modality Person Re-Identification With Shared-Specific Feature Transfer418
This paper proposes a novel cross-modality shared-specific feature transfer algorithm (termed cm-SSFT) to explore the potential of both the modality-shared information and themodality-specific characteristics to boost the reidentification performance.
- MiM-ISTD: Mamba-in-Mamba for Efficient Infrared Small-Target Detection227
Inspired by the recent basic model with linear complexity for long-distance modeling, Mamba, the potential of this state-space model (SSM) for ISTD tasks in terms of effectiveness and efficiency is explored.
- Demo225
This paper designs and implements Medusa, a novel programming framework for crowd-sensing that provides high-level abstractions for specifying the steps required to complete a crowd-Sensing task, and employs a distributed runtime system that coordinates the execution of these tasks between smartphones and a cluster on the cloud.
- Semantics Disentangling for Text-To-Image Generation210
A novel photo-realistic text-to-image generation model that implicitly disentangles semantics to both fulfill the high- level semantic consistency and low-level semantic diversity and a visual-semantic embedding strategy by semantic-conditioned batch normalization to find diverse low- level semantics.
- Fabrication of a pH-Responsively Controlled-Release Pesticide Using an Attapulgite-Based Hydrogel179
The PA-CA hydrogel possessed a benign biocompatibility on Escherichia coli and foxtail millet, showing a high biosafety, and provides a promising approach to pH-responsively controlled release chlorpyrifos.
- 155
- 149
- Online multi-object tracking with unsupervised re-identification learning and occlusion estimation111
This study shows that, when applied to state-of-the-art MOT methods, the proposed unsupervised re-Identification learning is comparable to supervised re-identification learning, and the tracking performance is further improved by the proposed occlusion estimation module.
- AResU-Net: Attention Residual U-Net for Brain Tumor Segmentation101
An end-to-end 2D brain tumor segmentation network, i.e., attention residual U- net (AResU-Net), which simultaneously embeds attention mechanism and residual units into U-Net for the further performance improvement of brain tumor segmentsation.
- Joint Color-irrelevant Consistency Learning and Identity-aware Modality Adaptation for Visible-infrared Cross Modality Person Re-identification98
This work proposes a novel approach to VI-ReID that learns the color-irrelevant features through the CICL and aligns the identity-level feature distributions by the Identity-aware modality adaptation (IAMA), which is integrated into a joint learning framework and can promote each other.
- Toward automatic quantification of knee osteoarthritis severity using improved Faster R-CNN97
The proposed end-to-end fully automatic model which is computationally efficient has the potential to achieve the real automatic diagnosis of knee OA and be used as computer-aided diagnosis tools in clinical applications.
- 94
- Context and Attribute Grounded Dense Captioning83
This work designs a novel end-to-end context and attribute grounded dense captioning framework consisting of a contextual visual mining module and a multi-level attribute grounded description generation module that incorporates an auxiliary supervision from hierarchical linguistic attributes to augment the distinctiveness of the learned captions.
- Breast cancer histopathological image classification using attention high‐order deep network80
A novel attention high‐order deep network (AHoNet) is introduced by simultaneously embedding attention mechanism and high-order statistical representation into a residual convolutional network to capture more discriminant deep features for breast cancer pathological images.
- Infrared-Light-Responsive Controlled-Release Pesticide Using Hollow Carbon Microspheres@Polyethylene Glycol/α-Cyclodextrin Gel70
An infrared-light-responsive pesticide delivery system for the controlled release of imidacloprid by preparation of functional hollow carbon microspheres by providing a promising method to intentionally regulate pesticide release and enhance utilization efficiency.
- Obstacle detection of rail transit based on deep learning66
The Improved-YOLOv4 network based on deep learning obtained the mAP of 93% on NVIDIA Jetson AGX, which is more suitable for the obstacle detection of rail transit.
- 64
- Approximate Nearest Neighbor Search under Neural Similarity Metric for Large-Scale Recommendation58
This paper proposes a novel method to extend ANN search to arbitrary matching functions, e.g., a deep neural network, and proposes a pluggable adversarial training task to ensure the graph search with arbitrary matching function can achieve fairly high precision.
- 57
- 57
- Lightweight and Secure Three-Factor Authentication Scheme for Remote Patient Monitoring Using On-Body Wireless Networks53
The proposed lightweight and secure three-factor authentication scheme for remote patient monitoring using oBWNs adopts one-time hash chain technique to ensure forward secrecy, and the pseudonym identity method is employed to provide user anonymity and resist against desynchronization attack.
- Real-Time Online Multi-Object Tracking in Compressed Domain52
This work divides the frames into key and non-key frames and track objects in the compressed domain and proposes an appearance convolutional neural network which can be jointly trained with the detector to make data association more reliable.
- Ship detection in optical sensing images based on YOLOv551
The results show that YOLOv5 can be well applied in the field of ship detection and that convolutional neural networks have achieved good results in ship target detection and recognition.
- Efficient and privacy-preserving authentication scheme for wireless body area networks51
In the proposed scheme, the method of certificateless authentication based on identity-based cryptography is adopted, which makes it suitable for multi-server architecture without online third party participation and reduce the computation cost effectively.
- 50
- 49
- 49
- 48
- Electrical-Driven Release and Migration of Herbicide Using a Gel-Based Nanocomposite46
This work provides a low cost and promising approach to control Gly release, deliver Gly precisely, and improve utilization efficiency, which might have a high application value.
- 43
- 43
- 43
- 40
- 39
- WiOpen: A Robust Wi-Fi-Based Open-Set Gesture Recognition Framework38
WiOpen, a robust Wi-Fi-based open-set gesture recognition (OSGR) framework that addresses uncertainty induced by noise during data preprocessing by utilizing the channel state information (CSI) ratio and designs the OSGR network based on an uncertainty quantification method, successfully realizing OSGR.
- DASOT: A Unified Framework Integrating Data Association and Single Object Tracking for Online Multi-Object Tracking38
An efficient two-stage tracking method in the SOT branch, which utilizes the merits of correlation features and can simultaneously track all the existing targets within one forward propagation, is proposed.
- ReSup: Reliable Label Noise Suppression for Facial Expression Recognition37
This paper proposes a more reliable noise-label suppression method called ReSup, which makes the noise data decision by modeling the distribution of noise and clean labels simultaneously according to the disagreement between the prediction and the target.
- Using multi-label classification to improve object detection36
A novel multi-task framework for object detection that uses multi-label classification as an auxiliary task to improve object detection, and can be trained and tested end-to-end and does not require any extra annotation.
- 36
- 32
- 32
- 31
- TDCA-Net: Time-Domain Channel Attention Network for Depression Detection30
An end-to-end time-domain channel attention network (TDCA-Net) for depression detection that directly models time-domain speech signals based on dilated convolution block and employs the efficient channel attention (ECA) module to model dependencies of channels and improve the of the model to information related to depression.
- 30
- Buffer-Aware Resource Allocation Scheme With Energy Efficiency and QoS Effectiveness in Wireless Body Area Networks29
A buffer-aware sensor evaluation method with low complexity is introduced to the resource allocation scheme to evaluate the sensor state in real time and then decide when applying for the resource re-allocation by the hub for further improving both the short-term and the long-term QoS performance.
- I Know Your Keyboard Input28
A robust side-channel attack scheme to infer keystrokes on the surrounding keyboard, leveraging the smart devices' microphones, and extract two robust features from the acoustic signals to alleviate the impact of various victims and keyboards is proposed.
- 28
- 27
- TACR-Net: Editing on Deep Video and Voice Portraits27
A novel deep learning framework, named Temporal-Refinement Autoregressive-Cascade Rendering Network (TACR-Net) for audio-driven dynamic talking face editing, which encodes facial expression blendshape based on the given acoustic features without separately training for special video.
- 26
- 26
- Fluid Dynamics-Inspired Network for Infrared Small Target Detection23
A novel research routine is explored by abstracting the movement of pixels in the ISTD process as the flow of fluid in fluid dynamics (FD), and a new Fluid Dynamics-Inspired Network (FDI-Net) is devised for ISTD.
- 23
- 22
- 22
- Towards Intrinsic Common Discriminative Features Learning for Face Forgery Detection Using Adversarial Learning21
This work proposes a novel method which utilizes adversarial learning to eliminate the negative effect of different forgery methods and facial identities, which helps classification network to learn intrinsic common discriminative features for face forgery detection.
- 20
- Second-order multi-instance learning model for whole slide image classification20
A novel second-order multiple instances learning method (SoMIL) with an adaptive aggregator stacked by the attention mechanism and recurrent neural network for histopathological image classification and an efficient channel attention mechanism to adaptively highlight the most discriminative instance features.
- 20
- Robust video object tracking via Bayesian model averaging-based feature fusion20
This work proposes an algorithm that is robust against occlusion, the presence of confusing colors, abrupt changes in the object features and changes in scale, and develops the algorithm within a Bayesian modeling framework.
- Exploiting Modality-Specific Features for Multi-Modal Manipulation Detection and Grounding19
This paper constructs a simple and novel transformer-based framework for multi-modal manipulation detection and grounding tasks and proposes an implicit manipulation query (IMQ) that adaptively aggregates global contextual cues within each modality using learnable queries, thereby improving the discovery of forged details.
- 19
- Semantic segmentation of high-resolution images19
This paper downsample the input image to a lower resolution and then obtain a low-resolution semantic segmentation image using state-of-the-art methods, which significantly reduces memory cost without reducing result quality.
- 19
- Multi-spectral Class Center Network for Face Manipulation Detection and Localization18
A novel Multi-Spectral Class Center Network (MSCCNet) is proposed for face manipulation detection and localization and a Multi-level Features Aggregation (MFA) module is proposed to employ more low-level forgery artifacts and structural textures.
- Chitosan‐based organic/inorganic composite engineered for UV light‐controlled smart pH‐responsive pesticide through in situ photo‐induced generation of acid18
The proposed in situ photo-induced generation of acid locally on the spots of applied pesticide circumvents the problem of acid contamination to nontargets, demonstrating higher efficiency and biocompatibility for the controlled delivery of acid-responsive pesticides and pest management.
- 17
- 17
- 15
- 15
- 13
- 13
- Repair of articular cartilage defect using adipose‐derived stem cell‐loaded scaffold derived from native cartilage extracellular matrix13
The results showed that the cartilage complex constructed by ADSCs and ACECM through tissue engineering successfully repaired thecartilage defect of the pig knee joint, indicating that ADSCs do not require in vitro induction and have the potential for chondrogenic differentiation in the environment around the knee joint.
- 13
- 13
- 13
- Learning Pixel-Level and Instance-Level Context-Aware Features for Pedestrian Detection in Crowds12
A novel context-aware feature learning method for detecting pedestrians in crowds, with the purpose of making better use of context information for dealing with occlusion, which achieves an efficient and lightweight detector that can be trained in an end-to-end fashion.
- 12
- MediaBroker: A pervasive computing infrastructure for adaptive transformation and sharing of stream data12
The EventWeb application demonstrates the utility and graceful scaling of MediaBroker for supporting pervasive computing applications and shows reasonable performance for selected streaming media-intensive applications.
- Extreme learning machine with autoencoding receptive fields for image classification11
The experimental results show that ELM-ARF effectively improves image classification accuracy by encoding the combined features that contain global attributes, which is used to extract global features.
- Flow-based Travel Plan via VANET11
A scheme of Flow-Based Travel plan of VANETs (FBTV), which can calculate most time-saving routes for vehicles via the real time traffic information of each road segments, and the traffic condition can be classified by the vehicles flow that is collected.
- 10
- CamLopa: A Hidden Wireless Camera Localization Framework via Signal Propagation Path Analysis9
CamLopa is introduced, a training-free wireless camera localization framework that operates with minimal activity space constraints using low-cost, commercial-off-the-shelf (COTS) devices and without the need for training.
- 9
- 9
- Two birds with one stone: Transforming and generating facial images with iterative GAN9
This work proposes to train GAN iteratively via regularizing the min-max process with an integrated loss, which includes not only the per-pixel loss but also the perceptual loss, and shows that the proposed iterative architecture can achieve both image generation and transformation.
- Looking into the world on Google Maps with view direction estimated photos9
The proposed system allows users to select the places they are interested in and then focus their exploration and walk freely through the map and view what can be seen along their eyes with the automatically returned precise set of photos of the target places/angles, by filtering out photos that are pointing to other directions.
- 9
- Fabrication of Tissue-Engineered Cartilage Using Decellularized Scaffolds and Chondrocytes8
It is found that ACECM-oriented scaffold is an ideal natural biomaterial for cartilage tissue engineering and has good biocompatibility.
- 8
- 8
- 8
- 8
- MODEL8
A new Moving Object Detection and Localization (MODEL) system, based on the smallscale fading of RF signal strength and independent from the salient characteristics of both the device and the sensor, which adopts an easily-realized Rolling-Window algorithm.
- 7
- ProMask: Probability mask representation for skeleton detection7
The ProMask is proposed, which is a novel skeleton detection model that includes the probability mask representation and vector router and considers that it will serve as a standard configuration for future skeleton detection, since it is reasonable, simple, and very effective.
- MultiRPN-DIDNet: Multiple RPNs and Distance-IoU Discriminative Network for Real-Time UAV Target Tracking7
Experimental results on the public datasets of UAV20L and DTB70 show that, compared with the state-of-the-art UAV trackers, the proposed MultiRPN-DIDNet can obtain better tracking performance with fewer region proposals and correction iterations.
- Remote Sensing Image Retrieval with Deep Features Encoding of Inception V4 and Largevis Dimensionality Reduction7
Experimental results on the datasets of RS19, UCM and RSSCN7 have demonstrated that the proposed remote sensing image retrieval method can obtain state-of-the-art retrieval performance.
- Polygonal approximation based on coarse-grained parallel genetic algorithm7
Experimental results, based on four benchmark curves and four real image curves extracted from the lake maps, show that the basic CGPGA outperforms the used genetic algorithm, and further the improved ICGPGA is more effective than the basicCGPGA, in terms of the quality ofbest solutions, the average solutions, and the variance of best solutions.
- 7
- 7
- S3: Smart Session Selection for Voice Communications in Next Generation Wireless Network7
A smart session selection (S3) scheme to meet this requirement, which enables users to select a best network connection, which consists of source and destination access network pair, to satisfy quality constraint and users' preference.
- Real-Time Steganography in Compressed Video7
An adaptive and large capacity steganography method applicable to compressed video with no need for full or even partial decompression and guided by several so-called A/S trees adaptively.
- Hidden WiFi Camera Localization via Signal Propagation Path Analysis6
CamLoPA is introduced, a system designed for the detection and localization of WiFi cameras that can accurately and swiftly detect and localize snooping WiFi cameras with minimal constraints.
- Evopose: A Recursive Transformer for 3D Human Pose Estimation with Kinematic Structure Priors6
A novel transformer-based model EvoPose is proposed to introduce the human body prior knowledge for 3D human pose estimation effectively and achieves a new state of the art on two most popular benchmarks, Human3.6M and MPI-INF-3DHP.
- 6
- 6
- Research on image classification method of features of combinatorial convolution6
The experimental results show that the low-dimensional image feature vectors generated by CCNet effectively reduce the classification error rate when the number of convolutional layers does not exceed 100 layers.
- PoseRL-Net: human pose analysis for motion training guided by robot vision5
PoseRL-Net not only improves pose estimation accuracy but also provides crucial support for intelligent decision-making and motion planning in robots operating in dynamic and complex scenarios, offering significant practical value for collaborative robotics.
- 5
- 5
- Cascaded Residual Density Network for Crowd Counting5
A novel Cascaded Residual Density Network (CRDNet) in a coarse-to-fine approach to generate the high-quality density map for crowd counting more accurately and reduces the errors of pixel-wise Euclidean loss by restricting the number of people in the local crowd areas.
- Scale Voting With Pyramidal Feature Fusion Network for Person Search5
This work proposes Pyramidal Feature Fusion Network that integrates the top-down mid-level features to provide multi-level feature outputs and designs a ranking algorithm named Scale Voting that votes among these ranking results from different feature levels to get the final ranking order.
- PPEDNet: Pyramid Pooling Encoder-Decoder Network for Real-Time Semantic Segmentation5
A pyramid pooling encoder-decoder network named PPEDNet is proposed for both better accuracy and faster processing speed and is evaluated on CamVid dataset, achieving 7.214% mIOU accuracy improvement while reducing 17.9% of the parameters compared with the state-of theart algorithm.
- 5
- 5
- 5
- 4
- Multi-spectral Class Center Network for Face Manipulation Localization4
It is observed that multi-frequency spectrum information is effective for identifying tampered regions of forgery images and a novel Multi-spectral Class Center Network (MSCCNet) is proposed for face manipulation localization.
- 4
- 4
- 4
- TCCF: Tracking Based on Convolutional Neural Network and Correlation Filters4
A novel algorithm for online visual object tracking which is divided into two separate parts, one of them is target location estimation and the other is target scale estimation which outperforms the state-of-the-art by great margin in terms of accuracy and robustness.
- 4
- A Structural Coupled-Layer Tracking Method Based on Correlation Filters4
A coupled-layer tracking method based on correlation filters that can not only have the advantages of existing correlation filter trackers, and the ability to handle partial occlusion well due to part-based strategy, but also preserve object structure.
- A Novel Research to Artificial Bandwidth Extension Based on Deep BLSTM Recurrent Neural Networks and Exemplar-Based Sparse Representation4
Experimental results demonstrate that the proposed framework can achieve significant improvements in both objective and subjective measures over the different baseline methods.
- 4
- 4
- Human Activity Recognition Using Smart-Phone Sensors4
A novel feature learning method for activity recognition based on entropy is proposed and an activity recognition model based on multi-class AdaBoost is constructed and shown to be effective.
- 4
- 4
- 4
- 4
- 4
- Using Temporal Semantics for Live Media Stream Queries4
This work investigates the modeling issues and introduces the query processing techniques of a live media stream management system (MedSMan), including media capturing, automatic feature generating, declaration and query languages, temporal stream operators and querying algorithms, and presents experimental results to show the performance of the prototype.
- 4
- 3
- A method framework of cruciate ligaments segmentation and reconstruction from MRI images3
The effectiveness of the proposed superpixel-based spectral clustering for cruciate ligament segmentation and its potential for increasing the segmentation accuracy and speed are shown, offering significant advantages over manual segmentation by reducing time and expertise.
- 3
- 3
- 3
- Affinity-Aware Relation Network for Oriented Object Detection in Aerial Images3
A novel and accessible Affinity-Aware Relation Network based on two-stage detectors based on two-stage detectors that achieves notable AP improvements as well as state-of-the-art performances on DOTA, HRSC2016 and UCAS-AOD datasets.
- Pose-Controllable 3D Facial Animation Synthesis using Hierarchical Audio-Vertex Attention3
A novel pose-controllable 3D facial animation synthesis method is proposed by utilizing hierarchical audio-vertex attention by utilizing the 2D talking face technique and results indicate that the proposed method can produce more realistic facial expressions and head posture movements.
- 3
- Consistent Relative Confidence and Label-Free Model Selection for Convolutional Neural Networks3
This paper proposes an approach to CNN model selection using only unlabeled data and develops this method based on a principle termed consistent relative confidence, which demonstrates the effectiveness and efficiency of this method.
- 3
- 3
- 3
- 3
- On relationship formation in heterogeneous information networks: An inferring method based on multilabel learning3
The experimental results show that by using heterogeneous information in a supervised MLL setting, MULRP achieves better performance in comparison to several baseline binary classification methods and a state‐of‐art relationship prediction method.
- 3
- Deep Scale Feature for Visual Tracking3
This paper proposes deep scale feature and an algorithm for robust visual tracking, inspired by correlation filter trackers, and proposes an online update algorithm to make the authors' tracker consistent with changing object in tracking video.
- 3
- A non-rigid registration method for cerebral DSA images based on forward and inverse stretching – avoiding bilinear interpolation3
A non-rigid registration method based on stretching transformation is presented in this paper, which is effective to maintain the topological relationships of the gray value before and after the image deformation.
- 3
- Resilient Recursive Routing in Communication Networks3
After a brief review of conventional approaches to shortest path routing an alternative algorithm that abstracts a network graph into a logical tree is introduced and can be used to analyze network graphs for resiliency, as well as become the basis of a new routing methodology.
- Performance Analysis of a Mobile Communication Network: Unidirectional Tandem Case with Phase Type Service3
An approximate method is developed to obtain main performances of systems such as the loss probabilities of handover calls and fresh calls, and the expected number of occupied channels in each node, based on decomposing the N cells in tandem into N−1 pairs of cells with overlap.
- 2
- Training-free Open-Vocabulary Semantic Segmentation via Diverse Prototype Construction and Sub-region Matching2
A novel training-free framework for OVSS that constructs diverse prototypes and performs fine-grained sub-region matching and leverages Large Language Models to guide image generation by descriptions of different attributes of categories and employs coarse-fine clustering to obtain diverse and robust part-level prototypes in the construction stage.
- 2
- 2
- 2
- Multimodality pedestrian detection based on YOLOv52
A pedestrian detection method based on YoloV5 RGB and thermal image data fusion that uses two yolov5 branch structures to learn the features of RGB and Thermal image data respectively, and finally uses multimodal features for fusion detection.
- An Easy Way to Deploy the Re-ID System on Intelligent City: Cross-Domain Few-Shot Person Reidentification2
A reweighting instance method based on influence function (ReWIF) to guide the training procedure of the re-ID model, which can effectively address the domain bias of different datasets and the absence of labeled data on the target dataset, achieving state-of-the-art performance.
- 2
- 2
- Fast Video Stitching for Aerially Captured HD Videos2
This work proposes a real-time video stitching system which is capable to stitch high definition (HD) videos captured by mobile aerial devices and uses scale invariant information to speed up the feature point extraction.
- Performance Analysis of IEEE 802.15.6 MAC Protocol in WBAN with Energy Harvesting Nodes2
A Discrete-Time Markov Chain (DTMC) model is built to analyze the IEEE 802.15.6 Medium Access Control (MAC) protocol under saturation condition in the WBAN with energy harvesting devices, finding that a better energy harvesting state will make the system perform better.
- Two Birds with One Stone: Iteratively Learn Facial Attributes with GANs.2
This paper proposes to train GAN iteratively via regularizing the minmax process with an integrated loss, which includes not only the per-pixel loss but also the perceptual loss, to argue that the perceptual information benefits the output of a high-quality image, while preserving the identity information.
- 2
- 2
- On the tag localization of web video2
A benchmark dataset for tag localization using multiple instance learning approach, which contains about 4,000 videos collected from YouTube portal by issuing 50 concepts as queries and carefully annotate the time durations when the tag appears in the video and also label the spatial location of object with mask in frames for object related tag.
- 2
- 2
- 1
- 1
- 1
- 1
- 1
- Sequence-to-subsequence learning with 1D-CNN for behind-the-meter solar generation disaggregation1
A data-driven approach using a sequence-to-subsequence (Seq2subseq) PV decomposition model based on the one-dimensional convolutional neural network (1D-CNN) that automatically extracts temporal features from net metered sequences and outputs behind-the-meter PV generation using a sliding window.
- 1
- 1
- 1
- Accurate and Time-saving Deepfake Detection in Multi-face Scenarios Using Combined Features1
This paper proposes a model that can perform accurate and time-saving Deepfake detection in multi-face scenarios and fuse different levels of features to improve the performance of the model and use single-face data to aid the training of the multi- face data.
- 1
- 1
- 1
- Research on rotor performance of air-bearing based on fluid-solid coupling1
To investigate whether the motorized spindle meets its technical specifications under working conditions, it is critical to analyze the performance of its rotor to verify whether the axial runout requirement can be satisfied under the working conditions of the spindle.
- 1
- Visual Tracking by Deep Discriminative Map1
A visual tracking algorithm based on deep discriminative map that guides the tracking algorithm by estimating the object’s size and shape, and whether it is proper to gather training samples is proposed.
- Binary "proximity patches motion" descriptor for action recognition in videos1
Experimental results on the Weizmann and KTH datasets show that the proposed method not only requires a small amount of memory, but is also faster and achieves competitive accuracy when compared to the state-of-the-art spatio-temporal binary descriptors.
- Investigating Efficient Feature Representation Methods and Training Objective for BLSTM-Based Phone Duration Prediction1
Perceptual tests indicate the proposed methods could improve the naturalness of the synthetic speech, which benefits from the proposed feature representation methods which could capture more precise contextual features, and the proposed training objective function could tackle the over-averaged problem for the generated phone durations.
- 1
- 1
- 1
- Rapid performance evaluation for ideal FROC and AFROC observers1
The proposed 3 reasonable assumptions under which the ideal observers reduce approximately to a particular form of a scan-statistic observer can be evaluated far more rapidly albeit with slight error than that of the originally proposed ideal observer.
- Study of Railway Bureau Dispatch Query System Based on Java Web Start1
The railway bureau real-time train diagram query system is employed as the example to elaborate the realization method and the reasonable basics are provided to the organization of railway bureau transportation activities.
- An optical measuring method of high temperature clinker layer's thickness1
A method based on CCD offering the advantages of high sensitivity, no electromagnetic interference and non-contact in order to measure the thickness of the cooler precisely in high temperature and dust atrocious condition online with real-time and reliable parameter for the control system is presented.
- –
- –
- –
- High-resolution image signal processing (ISP) technology based on stitching–
This paper introduces an image signal processing technology for high-ISP block processing and stitching method (IBPS method), demonstrating that this approach allows for the use of low-power hardware to access and process high-resolution images, thus meeting the requirements of transmission line monitoring.
- –
- –
- –
- –
- Spiking ST-former: enhancing spatio-temporal modeling in spiking transformers via integrated self-attention mechanisms–
Experimental results on both static and neuromorphic datasets demonstrate that STSA consistently outperforms spatial-only attention mechanisms and achieves competitive performance, highlighting the potential of STSA in enhancing SNN performance and advancing the development of spiking transformers.
- –
- –
- –
- –
- Lightweight Detection Algorithm for Air-to-Ground Threat Object–
An efficient lightweight detection algorithm is proposed, built upon the YOLOv7-tiny framework, to address the issue of large network parameters, lightweight modules ELAN-GS and Slim-Neck are employed and the NWD metric and NWD loss function are introduced.
- –
- –
- –
- –
- –
- –
- –
- –
- Enhanced Video Segmentation with Object Tracking–
This work uses tracking techniques to improve the performance of video object segmentation and tracking simultaneously in a unified framework and rejects outliers in the segmentation mask caused by background complexities, such as similar object inference or background clutter issues.
- A computer assisted automatic grenade throw training system with simple digital cameras–
A real-time computer assisted grenade throwing training system by simple digital camera and low-cost computational methods that may provide technical support for the modern military throwing training and may change the traditional manual measuring mode for grenade throwing.
- –
- Vison-Based Traffic Flow Prediction Using Dynamic Texture Model and Gaussian Process–
The results show that the proposed method beats a benchmark solution based on Gaussian mixture model, and has the potential to be developed into qualified and practical solutions to real-time TFP.
- –
- –
- –
- A Robust Occlusion Judgment Scheme for Target Tracking Under the Framework of Particle Filter–
A corner matching method is utilized to improve the accuracy of target position and velocity estimation due to structure information, thus obtain a more accurate weight value which can reflect the real target’s status.
- –
- Optimum Dynamic Shifting Opportunity Research of 3 Parameters Shifting for High Power AT–
Taking the AT of SUVs SX2300 for example, this paper builds shift schedule based on three parameters, and focuses on the method of calculating shift opportunity under dynamic property situation.
- Design of multimode inertial component data acquisition system based on SOPC–
Compared with the traditional measuring equipments, the multimode inertial component data measuring system has higher reliability and stability, which greatly promotes measurement efficiency and saves cost.
- Research of x-ray automatic image mosaic method–
The method of grayscale cutting pseudo-color enhancement was firstly used to complete the mapping transformation from gray to the pseudo- color, and to extract SIFT features from the images, and by making use of a similar measure of NCC (normalized cross correlation - Normalized cross-correlation) the method of RANSAC was used to exclude the pseudofeature points right in order tocomplete the exact match of feature points.
- A Monitoring Model of Unusual Transactions Based on Complex Event Processing–
A complex event processing technology (CEP) is introduced, a method of real-time monitoring for "big data", establishes a monitoring model of unusual transactions, designs and realizes a system based on this model.
- –
- –
- –
- –
- A Novel Message Signature Scheme for Sparse Mobile Ad Hoc Network–
Effective reduction of the key length, signature length and computation, and thus the network communication bandwidth and node memory and computation makes this scheme more suitable for sparse mobile Ad Hoc network.
- –
- –
- Weighted cooperative spectrum sensing technology based on data fusion–
Computer simulation results, show that weighted cooperative sensing method overcomes the disad- vantages of conventional Cooperative sensing method and outperforms the conventional method.
- A Microscope Traffic Parameter Extraction Method by Vehicle Tracking–
First, the vehicles are tracked by dynamic image processing techniques; then, a camera model is presented to obtain traffic parameters, and methods to obtain the vehicle position, estimate instantaneous vehiclespeed and some other traffic parameters are presented.
- –
- Research of correspondence points matching based on multi-scale wavelet transformation–
The complexity of correspondence points matching is decreased and the matching mistake is reduced using the area-based matching method coupled with feature-based method.
- –
- –
- –
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-14. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.