Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks392 from public data
- 3,666
- Bayesian CP Factorization of Incomplete Tensors with Automatic Rank Determination613
The method is characterized as a tuning parameter-free approach, which can effectively infer underlying multilinear factors with a low-rank constraint, while also providing predictive distributions over missing entries, which outperforms state-of-the-art approaches for both tensor factorization and tensor completion in terms of predictive performance.
- Dynamic visual attention: searching for coding length increments489
A dynamic visual attention model based on the rarity of features is proposed and the Incremental Coding Length (ICL) is introduced to measure the perspective entropy gain of each feature to maximize the entropy of the sampled visual features.
- Evo-ViT: Slow-Fast Token Evolution for Dynamic Vision Transformer317
Evo-ViT is presented, a self-motivated slow-fast token evolution approach for vision transformers that can accelerate vanilla transformers of both flat and deep-narrow structures from the very beginning of the training process.
- Credit Card Fraud Detection Using Convolutional Neural Networks310
A CNN-based fraud detection framework, to capture the intrinsic patterns of fraud behaviors learned from labeled data is proposed, on which a convolutional neural network is applied to identify a set of latent patterns for each sample.
- Leaf image based cucumber disease recognition using sparse representation classification291
This work proposes a novel cucumber disease recognition approach which consists of segmenting diseased leaf images by K-means clustering, extracting shape and color features from lesion information, and classifying disease leaf images using sparse representation (SR).
- DoveNet: Deep Image Harmonization via Domain Verification273
This work contributes an image harmonization dataset iHarmony4 by generating synthesized composite images based on COCO (resp., Adobe5k, Flickr, day2night) dataset, leading to the HCOCO sub-dataset, and proposes a new deep image harmonized method DoveNet using a novel domain verification discriminator.
- 269
- 259
- 258
- Bayesian Robust Tensor Factorization for Incomplete Multiway Data239
A generative model for robust tensor factorization in the presence of both missing data and outliers that can discover the groundtruth of CP rank and automatically adapt the sparsity inducing priors to various types of outliers is proposed.
- 217
- Taming the Power of Diffusion Models for High-Quality Virtual Try-On with Appearance Flow187
The approach, namely Diffusion-based Conditional Inpainting for Virtual Try-ON(DCI-VTON), effectively utilizes the power of the diffusion model, and the incorporation of the warping module helps to produce high-quality and realistic virtual try-on results.
- Feature learning from incomplete EEG with denoising autoencoder163
The Lomb-Scargle periodogram is employed to estimate the spectral power from incomplete EEG (after removing only parts contaminated by artefacts), and Denoising Autoencoder (DAE) for learning and the results show that the method can successfully decode incomplete EEG to good effect.
- MindFinder149
By scaling up the database to more than two million images, MindFinder not only helps users to easily present whatever they are imagining, but also has the potential to retrieve the most desired images in their mind.
- 137
- Neurophysiological substrates of stroke patients with motor imagery-based brain-computer interface training132
It is demonstrated that MI-based BCI training can enhance the motor function of the upper extremity for stroke patients by inducing the optimal cerebral motor functional reorganization.
- Spatio-Temporal Attention-Based Neural Network for Credit Card Fraud Detection126
The result shows that STAN performs better than other state-of-the-art baselines in both AUC and precision-recall curves and demonstrates the effectiveness of the proposed method in both detecting suspicious transactions and mining fraud patterns.
- 125
- High-Resolution Image Harmonization via Collaborative Dual Transformations120
This work proposes a high-resolution image harmonization network with Collaborative Dual Transformation (CDTNet) to combine pixel- to-pixel transformation and RGB-to-RGB transformation coherently in an end-to -end network.
- End-to-End Video Object Detection with Spatial-Temporal Transformers116
TransVOD is presented, an end-to-end video object detection model based on a spatial-temporal Transformer architecture that streamline the pipeline of VOD, effectively removing the need for many hand-crafted components for feature aggregation, e.g., optical flow, recurrent neural networks, relation networks.
- DESIGN OF ASSISTIVE WHEELCHAIR SYSTEM DIRECTLY STEERED BY HUMAN THOUGHTS114
An electroencephalogram (EEG)-based wheelchair which can be steered by users' own thoughts without any other involvements implies that people can steer wheelchair only by their thoughts, and may have a potential perspective in daily application for disabled people.
- Bargainnet: Background-Guided Domain Translation for Image Harmonization109
This work forms image harmonization task as background-guided domain translation, which is regulated by well-tailored triplet losses, and uses a domain code extractor to capture the background domain information to guide the foreground harmonization.
- The N-terminal Domain of NPC1L1 Protein Binds Cholesterol and Plays Essential Roles in Cholesterol Uptake108
It is demonstrated that plasma membrane-localized NPC1L1 binds exogenous cholesterol via its NTD, and facilitates the formation of NPC1 L1-flotillin-cholesterol membrane microdomains that are then internalized into cells through the clathrin-AP2 pathway.
- Knowledge Graph-based Event Embedding Framework for Financial Quantitative Investments106
A knowledge graph-based event embedding framework for quantitative investments that first extracts structured events from raw texts, and construct the knowledge graph with the mentioned entities and relations simultaneously, and leverages a joint model to merge theknowledge graph information into the objective function of anevent embedding learning model.
- XYLayoutLM: Towards Layout-Aware Multimodal Networks For Visually-Rich Document Understanding103
A robust layout-aware multimodal network named XYLayoutLM is proposed to capture and leverage rich layout information from proper reading orders produced by the Augmented XY Cut and achieves competitive results on document understanding tasks.
- F2GAN: Fusing-and-Filling GAN for Few-shot Image Generation100
A Fusing-and-Filling Generative Adversarial Network (F2GAN) is proposed to generate realistic and diverse images for a new category with only a few images to ensure the diversity of generated images by a mode seeking loss and an interpolation regression loss.
- Data Analytics for Environmental Science and Engineering Research95
An overview of data analytics frameworks suitable for various Environmental Science and Engineering research applications is provided and a path to advance incorporation of data Analytics approaches in ESE research and application is proposed.
- 84
- The Prevalence of Post-Traumatic Stress Disorder in the General Population during the COVID-19 Pandemic: A Systematic Review and Single-Arm Meta-Analysis83
The COVID-19 pandemic brought certain mental pain to general population, leading to a rise in the incidence of PTSD in a short time, and this systematic review and single-arm meta-analysis found that PTSD in the general population was 15% during the epidemic.
- From Representation to Reasoning: Towards both Evidence and Commonsense Reasoning for Video Question-Answering81
Through extensive experiments on existing VideoQA methods, it is found that the state-of-the-art methods are strong in descriptions but weak in reasoning, and it is hoped that Causal-VidQA can guide the research of video understanding from representation learning to deeper reasoning.
- Matchinggan: Matching-Based Few-Shot Image Generation76
This work proposes matching-based Generative Adversarial Network for few-shot generation, which includes a matching generator and a matching discriminator that extends conventional GAN discriminator by matching the feature of generated image with the fused feature of conditional images.
- 76
- Multi-shot Pedestrian Re-identification via Sequential Decision Making71
In three open benchmarks, the proposed interpretable reinforcement learning based approach to multi-shot pedestrian re-identification problem are competitive with the state-of-the-art methods while only using 3% to 6% images.
- Free Hand-Drawn Sketch Segmentation67
A real-time solution is presented to automatically segment a user's sketch during his/her drawing using a graph-based sketch segmentation algorithm and a semantic-based approach to simulate the past experience in the perceptual system by leveraging a web-scale clipart database.
- Risk Assessment for Networked-guarantee Loans Using High-order Graph Attention Representation65
A high-order graph attention representation method (HGAR) to learn the embedding of guarantee networks and proves that the complexity is nearlinear with the number of edges, which could scale to large datasets.
- Sketch-based image retrieval on a large scale database65
A simple and effective sketch-based algorithm for large scale image retrieval which could increase the retrieval rate dramatically and first extract orientation features and then organize them in a hierarchal way to generate global-to-local features.
- The Anxiety Status of Chinese Medical Workers During the Epidemic of COVID-19: A Meta-Analysis64
The anxiety level of Chinese medical workers has increased significantly during the epidemic of COVID-19 and was significantly higher than that of the national norm in each study.
- Video Semantic Segmentation via Sparse Temporal Transformer63
This work proposes a novel Sparse Temporal Transformer (STT) to bridge temporal relation among video frames adaptively, which is also equipped with query selection and key selection, and achieves the state-of-the-art segmentation accuracy and temporal consistency with comparable inference speed.
- Disentangled Information Bottleneck61
Disentangled Information Bottleneck (DisenIB) is introduced that is consistent on compressing source maximally without target prediction performance loss (maximum compression), and performs well in terms of generalization, robustness to adversarial attack, out-of-distribution detection, and supervised disentangling.
- A Boosting-Based Spatial-Spectral Model for Stroke Patients’ EEG Analysis in Rehabilitation Training61
- 61
- A tensor-based scheme for stroke patients’ motor imagery EEG analysis in BCI-FES rehabilitation training60
A tensor-based scheme to detect motor imagery EEG patterns in spatial-spectral-temporal domain directly from multidimensional EEG constructed by wavelet transform method, preserving the structural information of the multi-channel time-varying EEG.
- Blind Deconvolution of Dynamical Systems:A State-Space Approach60
A new method is proposed: the two-stage approach to blind deconvolution in the state-space framework, and how to extend the information backpropagation approach to the nonlinear case is suggested.
- Image Cropping with Composition and Saliency Aware Aesthetic Score Map59
This paper proposes an interpretable image cropping model that can be used to localize aesthetically important regions in an image, which sheds light on the composition rules learned by the model.
- 54
- 50
- Image Composition Assessment with Saliency-augmented Multi-pattern Pooling49
A composition assessment network SAMP-Net with a novel Saliency-Augmented Multi-pattern Pooling (SAMP) module, which analyses visual layout from the perspectives of multiple composition patterns, and can perform more favorably than previous aesthetic assessment approaches.
- Arteriosclerosis Assessment Based on Single‐Point Fingertip Pulse Monitoring Using a Wearable Iontronic Sensor46
A hfPWV‐based model is established for arteriosclerosis evaluation with an accuracy comparable to that of existing clinical criteria, and the validity of the model is verified clinically.
- 44
- 43
- 42
- Hard Pixel Mining for Depth Privileged Semantic Segmentation42
This paper proposes a novel Loss Weight Module, which outputs a loss weight map by employing two depth-related measurements of hard pixels: Depth Prediction Error and Depth-aware Segmentation Error, and is applied to segmentation loss, with the goal of learning a more robust model by paying more attention to the hard pixels.
- A Phase Discrepancy Analysis of Object Motion42
A surprisingly simple algorithm to detect objects in such conditions by showing that the displacement of the foreground and the background can be represented by the phase change of Fourier spectra and the motion of background objects can be extracted by Phase Discrepancy in an efficient and robust way.
- Action-Aware Embedding Enhancement for Image-Text Retrieval41
Action-aware Memory-Enhanced embedding (AME) method for image-text retrieval is proposed, which aims to emphasize the action information when mapping the images and texts into a shared embedding space and integrates action prediction along with an action-aware memory bank to enrich the image and text features with action-similar text features.
- 41
- 40
- Multilinear Subspace Regression: An Orthogonal Tensor Decomposition Approach40
The proposed multilinear subspace regression model based on so called latent variable decomposition is introduced and is shown to be suitable for the prediction of multidimensional dependent data from multiddimensional independent data.
- 40
- 38
- Cardiology knowledge free ECG feature extraction using generalized tensor rank one discriminant analysis38
This study introduced a generalized tensor rank one discriminant analysis (GTR1DA), which involves considering the distribution of the data points near the classification boundary to calculate better projection tensors and achieves greater classification accuracy than other vector- and tensor-based methods.
- Query-adaptive shape topic mining for hand-drawn sketch recognition38
A query-adaptive shape topic model is proposed to mine object topics and shape topics related to the sketch, in which, multiple layers of information such as sketch, object, shape, image, and semantic labels are modeled in a generative process.
- 37
- 37
- A Dynamic Default Prediction Framework for Networked-guarantee Loans35
A dynamic default prediction framework (DDPF) is proposed, which preserves temporal network structures and loan behavior sequences in an end-to-end model and design a gated recursive and attention mechanism to integrate both the loan behavior and network information.
- Active training paradigm for motor imagery BCI35
This work proposed a new paradigm where subjects participated in training more actively than in the traditional paradigm, and this active training paradigm may generate better training samples with fewer inconsistent labels because it overcomes mistakes when subject’s motor imagination does not match the given cues.
- 35
- DeepMicroGen: a generative adversarial network-based method for longitudinal microbiome data imputation34
This work proposes DeepMicroGen, a bidirectional recurrent neural network-based GAN model, trained on the temporal relationship between the observations, to impute the missing microbiome samples in longitudinal studies, and outperforms standard baseline imputation methods, showing the lowest mean absolute error for both simulated and real datasets.
- 34
- 34
- Deep Image Harmonization in Dual Color Spaces32
This paper explores image harmonization in dual color spaces, which supplements entangled RGB features with disentangled L, a, b features to alleviate the workload in harmonization process.
- Multi-Attribute Transfer via Disentangled Representation32
A novel model formulating disentangled representations by projecting images to latent units, grouped feature channels of Convolutional Neural Network, to disassemble the information between different attributes to retain the diversity beyond the labels inside each image.
- Sketch-based Image Retrieval via Shape Words32
This paper proposes to leverage shape words descriptor for sketch-based image retrieval by generalizing the classic Chamfer Matching algorithm to address the shape words matching problem.
- 32
- Regularized tensor discriminant analysis for single trial EEG classification in BCI32
A regularized tensor discriminant analysis (RTDA) algorithm is proposed for a multi-way discriminative subspace extraction from tensor-represented EEG data, which improves the performance for EEG classification and is able to find the most significant channels for classification, and can be applied to channel selection in BCI.
- Approximate maximum likelihood source separation using the natural gradient32
A maximum likelihood approach to source separation in the case of overdetermined mixtures corrupted by additive white Gaussian noise is addressed and a natural gradient adaptation algorithm is derived which maximizes the corresponding approximate likelihood function.
- Activity Image-to-Video Retrieval by Disentangling Appearance and Motion31
This paper proposes a Motion-assisted Activity Proposal-based Image-to-Video Retrieval (MAP-IVR) approach to disentangle the video features into motion features and appearance features and obtain appearance features from the images.
- 31
- Comparative efficacy and adherence of telehealth cardiac rehabilitation interventions for patients with cardiovascular disease: A systematic review and network meta-analysis30
The results indicated that telehealth cardiac rehabilitation improved VO2 peak, 6-minute walking distance, moderate-to-vigorous intensity physical activity, and adherence, and combined two or more types of telehealth interventions was found to be effective.
- Painterly Image Harmonization using Diffusion Model30
A novel Painterly Harmonization stable Diffusion model (PHDiffusion), which includes a lightweight adaptive encoder and a Dual Encoder Fusion (DEF) module, which can stylize the foreground more sufficiently and simultaneously retain finer content.
- 29
- Amodal Instance Segmentation via Prior-Guided Expansion29
A prior-guided expansion framework is proposed, which builds on a two-stage segmentation model and performs box-level (resp., pixel-level) expansion for amodal box (resp, mask) prediction, by retrieving regression transformations from a memory bank of expansion prior.
- Risk Guarantee Prediction in Networked-Loans29
An attention-based graph neural network to encode the edges that preserve the financial status as well as network structures in dynamic loan networks and shows that DGANN could significantly improve the risk prediction accuracy compared with state-of-the-art baselines.
- A Note on Lewicki-Sejnowski Gradient for Learning Overcomplete Representations29
A brief and more rigorous mathematical proof for this gradient is provided and a more robust constrained Lewicki-Sejnowski gradient is proposed.
- 28
- Uncorrelated Multiway Discriminant Analysis for Motor Imagery EEG Classification28
A robust Tensor-based method is proposed for a multiway discriminative subspace extraction from tensor-represented EEG data, which performs well in motor imagery EEG classification without the prior neurophysiologic knowledge like channels configuration and active frequency bands.
- Development of a Novel Enzymatic Cycling Assay for Total Homocysteine28
Highly increased concentrations of tHCY are found in persons with homocystinuria, a rare genetic disorder most commonly caused by a deficiency of cystathionine β-synthase, and patients with chronic renal disease have high morbidity and mortality resulting from …
- Advances in Neural Networks - ISNN 201027
Introducing a new hobby for other people may inspire them to join with you, and one that the authors will refer to break the boredom in reading is choosing advances in neural networks as the reading material.
- 25
- Learning From Web Data With Self-Organizing Memory Module25
A novel method, capable of handling these two types of noises together, without the supervision of clean images in the training stage, is proposed, under the framework of multi-instance learning.
- Temporal and Spatial Features of Single-Trial EEG for Brain-Computer Interface25
A new framework of feature extraction for classification of hand movement imagery EEG is proposed using independent residual analysis (IRA) method and the CSP method to obtain the optimal spatial and temporal features with which to achieve the best classification rate.
- Age and gender differences in static and dynamic balance of Chinese preschool children24
It is revealed that static and dynamic balance stability in preschool-aged children was affected by gender and age, and age- and gender-specific balance performance reference values for preschool children across multiple methods are provided.
- From Pixel to Patch: Synthesize Context-Aware Features for Zero-Shot Semantic Segmentation24
A novel context-aware feature generation network (CaGNet) that can synthesize context- aware pixel-wise visual features for unseen categories based on category-level semantic representations and pixel- wise contextual information is proposed.
- GAIN24
The proposed Gradient Augmented Inpainting Network (GAIN), which uses image gradient information instead of edge information to facilitate image inpainting, is formulated, which outperforms state-of-the-art methods quantitatively and qualitatively.
- Flexible Component Analysis for Sparse, Smooth, Nonnegative Coding or Representation24
A set of local cost functions with flexible penalty and regularization terms whose simultaneous or sequential minimization via a projected gradient technique leads to simple Hebbian-like local algorithms that work well not only for an over-determined case but also (under some weak conditions) for an under-d determined case.
- 22
- 22
- 21
- A Proposal-Based Approach for Activity Image-to-Video Retrieval21
This paper proposes an Activity Proposal-based Image-to-Video Retrieval (APIVR) approach, which incorporates multi-instance learning into cross-modal retrieval framework to address the proposal noise issue and proposes geometry-aware triplet loss based on point- to-subspace distance to preserve the structural information of activity proposals.
- 21
- Modeling Similarities Among Multi-Dimensional Financial Time Series21
Experiments on stocks data of all constituent firms of the S&P 500 demonstrate the superior performance of the proposed tensor-based framework when compared with some state-of-the-art methods.
- Generalized optimal spatial filtering using a kernel approach with application to EEG classification21
A generalized CSP (GCSP) based on generalized singular value decomposition (GSVD) and kernel method is proposed which is able to find the nonlinear spatial filters which are formulated in the feature space defined by a nonlinear mapping through kernel functions.
- Color conceptualization21
This paper proposes a method to manipulate colors of an image based on a library of natural color images, which evolves several prototypes of color distribution of the library, which it calls "color concepts".
- Zero-shot sketch-based image retrieval with structure-aware asymmetric disentanglement20
This work proposes the STRucture-aware Asymmetric Disentanglement (STRAD) method, in which image features are disentangled into structure features and appearance features while sketch features are only projected to structure space.
- Fashion Sensitive Clothing Recommendation Using Hierarchical Collocation Model20
A novel clothing recommendation approach which is sensitive to the fashion trend is proposed, which is able to learn clothing collocation metric from multi-dimensional clothing information and ensures the fashion sensitivity of the recommended outfits.
- ECG Classification Using ICA Features and Support Vector Machines20
A statistical method to segment heartbeats from ECG signal as precisely as possible is proposed, and the combination of independent component analysis (ICA) features and temporal feature to describe multi-lead ECG signals is used.
- STC: Spatio-Temporal Contrastive Learning for Video Instance Segmentation19
This work presents a simple and efficient single-stage VIS framework based on the instance segmentation method CondInst by adding an extra tracking head and proposes a novel bi-directional spatio-temporal contrastive learning strategy for tracking embedding across frames.
- Image Denoising with Rectified Linear Units19
This paper adopts the rectified linear (ReL) function instead of the sigmoid function as the activation function of hidden layers to further enhance the ability of neural network on solving image denoising problem.
- 19
- 19
- Thumbnail Generation Based on Global Saliency19
This method crops an image into a smaller but more informative region in the thumbnail representation, and evaluates the statistical distribution of feature maps, and uses its coding length as a measurement for image cropping.
- 18
- 18
- 18
- Extracting hierarchical spatial and temporal features for human action recognition18
A dual-channel model is presented to decouple the spatial and temporal feature extraction to capture the complementary static form information from single frame and dynamic motion information from multi-frame differences in two separate channels.
- Multilinear and nonlinear generalizations of partial least squares: an overview of recent advances17
The PLS‐related methods including linear, multilinear, and nonlinear variants and the extension of CCA to tensor space are overviewed and several examples are given to compare these methods with respect to the regression and classification techniques.
- Deep Image Harmonization with Learnable Augmentation16
This work explores learnable augmentation to enrich the illumination diversity of small-scale datasets for better harmonization performance and designs a designed SYthetic COmposite Network that takes in a real image with foreground mask and a random vector to learn suitable color transformation, which is applied to the foreground of this real image to produce a synthetic composite image.
- 16
- Multifactor sparse feature extraction using Convolutive Nonnegative Tucker Decomposition16
A flexible two stage algorithm for K-mode Convolutive Nonnegative Tucker Decomposition (K-CNTD) model is proposed using an alternating least square procedure and indicates that it is able to improve the speaker recognition performance especially in noisy conditions and has potential application on AD diagnosis.
- Overcomplete topographic independent component analysis16
A new method using both topographic and overcomplete representations of natural images, showing emergence of properties similar to those of complex cells in primary visual cortex (V1), can be considered as an extension of model in Hyvarinen et al.
- Few-shot Image Generation Using Discrete Content Representation15
This work makes the first attempt to adapt few-shot image translation method to few- shot image generation task, and model the autoregressive distribution of discrete content map conditioned on style vector, which can alleviate the incompatibility between content map and style vector.
- 15
- 15
- Age Classification System with ICA Based Local Facial Features15
An automatic age classification system that distinguishes kid faces from adult faces in real time by automatically locating, cropping and aligning faces from images and performing classification by SVM is proposed.
- Blind source estimation of FIR channels for binary sources: a grouping decision approach15
A novel grouping decision approach for blind source estimation of FIR channels with binary sources using a new deterministic algorithm based on grouping and decision to recover the source up to a delay.
- Slice Oriented Tensor Decomposition of EEG Data for Feature Extraction in Space, Frequency and Time Domains14
A novel tensor decomposition model of SOD (slice oriented decomposition) is applied to extract slice features from the multichannel time-frequency representation of EEG signals measured for MI (motor imagery) tasks in application to BCI (brain computer interface).
- 13
- 13
- 13
- Real Time Crowd Counting with Human Detection and Human Tracking13
Experiments on a real-world video show the proposed method for estimating the number of people in crowded surveillance scenes can give an accurate estimation in real time.
- 13
- 12
- An Iterative Method for Classifying Stroke Subjects’ Motor Imagery EEG Data in the BCI-FES Rehabilitation Training System12
An adaptive CSP method is proposed to deal with unknown irregular patterns in motor imagery signals of stroke patients and is applied on the EEG datasets of several stroke subjects comparing with traditional CSP-SVM.
- 12
- The Efficiency of Respiratory Exercises in Rehabilitation of Low Back Pain: A Systematic Review and Meta-Analysis11
An exercise program with 30 to 50 minutes, 3 to 5 times per week, and >4 weeks of breathing resistance exercise program as the most effective for treating LBP is recommended.
- Generalizing predictions to unseen sequencing profiles via deep generative models11
DeepBioGen is proposed, a sequencing profile augmentation procedure that characterizes visual patterns of sequencing profiles, generates realistic profiles based on a deep generative model capturing the patterns, and generalizes the subsequent classifiers that surpass the state-of-the-art method.
- 11
- Multi‐mode neural network for human action recognition11
This study proposes an effective framework that takes the advantage of deep learning on the static image feature extraction to tackle the video data and achieves a much better performance than the existing long short-term memory-based methods and consumes far fewer resources than theexisting 3D end-to-end models.
- A Comparative Analysis of Read Mapping and Indel Calling Pipelines for Next-Generation Sequencing Data11
This chapter evaluated the precision and recall of combinations of read mapping and indel calling software on calling short and longer indels with variable read coverage and read length on simulated data.
- Auditory Sparse Representation for Robust Speaker Recognition Based on Tensor Structure11
Experimental results demonstrate that the proposed nonnegative tensor principal component analysis (NTPCA) with sparse constraint can increase the recognition accuracy specifically in noisy environments.
- 10
- 10
- Sketch Recognition with Natural Correction and Editing10
Based on the proposed algorithms, a real-time sketch recognition system is built to recognize hand-drawn flowcharts and diagrams with flexible interactions.
- 10
- 9
- 9
- 9
- 9
- 9
- 9
- 9
- 9
- 9
- 9
- 9
- 8
- 8
- 8
- Visual analytics for loan guarantee network risk management.8
This is the first visual analytics tool to explore the guarantee network risks in a systematic manner and can be useful for the government and bank to monitor the default spread status and the temporal diffusion path analysis.
- 8
- 8
- 8
- 8
- 8
- 8
- 8
- 8
- 7
- 7
- 7
- 7
- 7
- Exploiting Motion Information from Unlabeled Videos for Static Image Action Recognition7
This paper integrates the above two strategies in a unified framework, which consists of Visual Representation Enhancement (VRE) module and Motion Representation Augmentation (MRA) module, and demonstrates the superiority of the framework based on four benchmark human action datasets with limited labeled data.
- 7
- A hybrid approach for risk assessment of loan guarantee network7
A hybrid feature representation is proposed, which is feeded into a gradient boosting model for credit risk assessment of guarantee network, and shows promising prediction performance on real-world data, which can be useful to both regulators and stakeholders.
- 7
- 7
- 6
- 6
- 6
- 6
- 6
- 6
- 6
- 5
- 5
- 5
- BHONEM: Binary High-Order Network Embedding Methods for Networked-Guarantee Loans5
A binary higher-order network embedding method to learn the low-dimensional representations of a guarantee network and shows that this method outperforms other start-of-the-art algorithms for both classification accuracy and robustness, especially in the guarantee network.
- 5
- 5
- 5
- 5
- 5
- 5
- 5
- 5
- 5
- 5
- 5
- 5
- 4
- Rethinking Classifier Re-Training in Long-Tailed Recognition: A Simple Logits Retargeting Approach4
This study revisit classifier re-training methods based on a unified feature representation and proposes a new metric called Logits Magnitude as a superior measure of model performance, replacing the commonly used Weight Norm.
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- Scene Gist: A Holistic Generative Model of Natural Image4
This paper proposes a novel generative model for natural image representation and scene classification that is decomposed with learned holistic basis called scene gist components, a global and adaptive image descriptor, generatively including most essential information related to visual perception.
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- Hierarchical Semantic Classification and Attribute Relations Analysis with Clothing Region Detection3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 2
- 2
- 2
- 2
- Towards Automatic Curation of Antibiotic Resistance Genes via Statement Extraction from Scientific Papers: A Benchmark Dataset and Models2
To the best of the knowledge, this work is the first to leverage natural language processing techniques to curate all validated ARGs from scientific papers and exploit three state-of-the-art neural architectures based on pre-trained language models and prompt tuning, and further ensemble them to attain the highest 77.0% F-score.
- 2
- 2
- 2
- Learning from Web Data with Memory Module2
A novel method, capable of handling these two types of noises together, without the supervision of clean images in the training stage, is proposed, under the framework of multi-instance learning.
- ProtoNet: Learning from Web Data with Memory2
The novel ProtoNet is proposed, which is capable of handling these two types of noises together, without the supervision of clean images in the training stage, and can be easily integrated into arbitrary CNN model.
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- A Novel Hierarchical Model of Attention: Maximizing Information Acquisition2
A novel attention model to produce saliency maps by generating information distributions on incoming images by automatically marks spots with large information amount as saliency, which ensures the system gains the maximum information acquisition through attending these spots.
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- A Hierarchial Model for Visual Perception1
Analysis of the perceptual manifold reveals that scene images which share similar perceptual similarities stay nearby in the manifold space, and the dimensions of the space could describe the spatial layout of scenes, which are like the degree of naturalness, openness supervised trained in this model.
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- –
- –
- –
- HM-NVS: Hierarchical Multi-Modal Novel View Synthesis\\with Uncertainty-Aware Progressive Refinement–
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
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 2026-10-16. 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.