Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks340 from public data
- 1,164
- Relation-Shape Convolutional Neural Network for Point Cloud Analysis1,050
RS-CNN, namely, Relation-Shape Convolutional Neural Network, which extends regular grid CNN to irregular configuration for point cloud analysis, achieves the state of the arts and a hierarchical architecture can be developed to achieve contextual shape-aware learning for Point cloud analysis.
- Learning a Mahalanobis distance metric for data clustering and classification637
This paper considers a general problem of learning from pairwise constraints in the form of must-links and cannot-links, and aims to learn a Mahalanobis distance metric.
- 622
- AugFPN: Improving Multi-Scale Feature Learning for Object Detection588
This paper begins by first analyzing the design defects of feature pyramid in FPN, and then introduces a new feature pyramid architecture named AugFPN to address these problems, which consists of three components: Consistent Supervision, Residual Feature Augmentation, and Soft RoI Selection.
- 570
- 473
- Face Detection Based on Multi-Block LBP Representation447
This paper presents the use of a new set of distinctive rectangle features, called Multi-block Local Binary Patterns (MB-LBP), for face detection, which encodes rectangular regions' intensities by local binary pattern operator, and the resulting binary patterns can describe diverse local structures of images.
- 437
- Trace ratio criterion for feature selection378
A novel algorithm is proposed to efficiently find the global optimal feature subset such that the subset-level score is maximized, and extensive experiments demonstrate the effectiveness of the proposed algorithm in comparison with the traditional methods for feature selection.
- 326
- Spatio-Temporal Graph Structure Learning for Traffic Forecasting314
A novel framework named Structure Learning Convolution (SLC) is proposed that enables to extend the traditional convolutional neural network (CNN) to graph domains and learn the graph structure for traffic forecasting and significantly outperforms the state-of-the-art ones.
- DensePoint: Learning Densely Contextual Representation for Efficient Point Cloud Processing300
DensePoint is proposed, a general architecture to learn densely contextual representation for point cloud processing that extends regular grid CNN to irregular point configuration by generalizing a convolution operator, which holds the permutation invariance of points, and achieves efficient inductive learning of local patterns.
- Semantic labeling in very high resolution images via a self-cascaded convolutional neural network290
A novel deep model with convolutional neural networks (CNNs), i.e., an end-to-end self-cascaded network (ScasNet), for confusing manmade objects and fine-structured objects, ScasNet improves the labeling coherence with sequential global- to-local contexts aggregation.
- Spectral Unmixing via Data-Guided Sparsity276
This paper proposes a novel sparsity-based method by learning a data-guided map (DgMap) to describe the individual mixed level of each pixel and applies the ℓp (0 <; p <; 1) constraint in an adaptive manner.
- 263
- Data guided Sparse Method for Hyperspectral Unmixing.257
This paper proposes a novel sparsity-based method by learning a data-guided map (DgMap) to describe the individual mixed level of each pixel and applies the ℓp (0 <; p <; 1) constraint in an adaptive manner.
- 243
- 222
- Gated Convolutional Neural Network for Semantic Segmentation in High-Resolution Images201
A gated convolutional neural network is proposed that achieves competitive segmentation accuracy on the public ISPRS 2D Semantic Labeling benchmark, which is challenging for segmentation by only using the RGB images.
- GSTNet: Global Spatial-Temporal Network for Traffic Flow Prediction189
A Global Spatial-Temporal Network (GSTNet) is proposed, which consists of several layers of spatial-temporal blocks, which can simultaneously extract the dynamic temporal dependencies and the global spatial correlations.
- 173
- A unified framework for semi-supervised dimensionality reduction173
This paper proposes a semi-supervised dimensionality reduction framework, which can efficiently handle the unlabeled data and can significantly improve the accuracy rates of the corresponding supervised and unsupervised approaches.
- Semi-supervised orthogonal discriminant analysis via label propagation154
This paper proposes a novel semi-supervised orthogonal discriminant analysis via label propagation that propagates the label information from the labeled data to the unlabeled data through a specially designed label propagation, and thus the distribution of the unl labeled data can be explored more effectively to learn a better subspace.
- 147
- A general graph-based semi-supervised learning with novel class discovery135
The core idea of the algorithm is to not only achieve the goal of semi-supervised learning, but also to discover the latent novel class in the data, which may be unlabeled by the user.
- 134
- 133
- 127
- 123
- 123
- 122
- Pro-Tuning: Unified Prompt Tuning for Vision Tasks120
Comprehensive experiments evidence that the proposed Pro-tuning outperforms fine-tuning on a broad range of vision tasks and scenarios, including image classification, class imbalance, image corruption, natural adversarial examples, and out-of-distribution generalization.
- 120
- 116
- 114
- Semisupervised Dimensionality Reduction and Classification Through Virtual Label Regression107
The proposed approach propagates the label information from labeled to unlabeled data with a well-designed mechanism of random walks, in which outliers are effectively detected and the obtained virtual labels of unlabeling data can be well encoded in a weighted regression model.
- 106
- You Only Search Once: Single Shot Neural Architecture Search via Direct Sparse Optimization103
The motivation behind DSO-NAS is to address the task in the view of model pruning, and it enjoys both advantages of differentiability and efficiency, therefore it can be directly applied to large datasets like ImageNet and tasks beyond classification.
- Neighborhood MinMax projections103
A new algorithm, Neighborhood MinMax Projections (NMMP), is proposed for supervised dimensionality reduction, and focuses only on the pairwise points where the two points are neighbors of each other, which has three significant advantages.
- 102
- Progressive Sparse Local Attention for Video Object Detection99
A novel module called Progressive Sparse Local Attention (PSLA) is proposed, which establishes the spatial correspondence between features across frames in a local region with progressively sparser stride and uses the correspondence to propagate features.
- 96
- A general kernelization framework for learning algorithms based on kernel PCA95
This paper gives a rigorous justification to reveal that under some mild conditions, the kernelization under this framework is equivalent to traditional kernel method, and shows that these mild conditions are usually satisfied in most of learning algorithms.
- A Performance Evaluation of Local Features for Image-Based 3D Reconstruction89
This paper performs a comprehensive and comparative evaluation of the state-of-the-art local features for the task of image-based 3D reconstruction and shows that binary features are competent to reconstruct scenes from controlled image sequences with only a fraction of processing time compared to using float type features.
- 84
- Robust Hyperspectral Unmixing With Correntropy-Based Metric83
This paper presents a robust model for unsupervised hyperspectral unmixing with the correntropy-based metric where the nonnegative constraints on both endmembers and abundances are imposed to keep physical significance.
- 79
- Extracting the optimal dimensionality for local tensor discriminant analysis77
This paper proposes an algorithm to extract the optimal dimensionality for local tensor discriminant analysis and results on a toy example and real-world data validate the effectiveness of the proposed method.
- Stitcher: Feedback-driven Data Provider for Object Detection76
Stitcher is presented, a feedback-driven data provider, which aims to train object detectors in a balanced way, and steadily improves performance by a large margin in all settings, especially for small objects, with nearly no additional computation in both training and testing stages.
- Learning Where to Focus for Efficient Video Object Detection70
A novel module called Learnable Spatio-Temporal Sampling (LSTS) has been proposed to learn semantic-level correspondences among adjacent frame features accurately and achieves state-of-the-art performance on the ImageNet VID dataset with less computational complexity and real-time speed.
- 70
- CAN69
A drug–drug synergy mediated by allosteric cross-talk in chromatin, whereby the binding of one drug alters the activity of the second, is described, which opens up possibilities for epigenetic targeting and suggests that allosterics modulation in nucleosomes may have biological relevance and potential for therapeutic interventions.
- 67
- Accurate urban road centerline extraction from VHR imagery via multiscale segmentation and tensor voting66
A fitting based algorithm is proposed to overcome the ineffectiveness of existing methods in the road intersections and achieve higher quantitative results, as well as more satisfactory visual performances by comparing with state-of-the-art methods.
- Orthogonal vs. uncorrelated least squares discriminant analysis for feature extraction66
It is shown that the orthogonal least squares discriminant analysis is an extension to the null space linear discriminative analysis, and the uncorrelated least squares discrimination analysis is exactly equivalent to the traditional linear discriminantAnalysis.
- 66
- 65
- 63
- Road Centerline Extraction via Semisupervised Segmentation and Multidirection Nonmaximum Suppression61
- 61
- 61
- Local and Global Regressive Mapping for Manifold Learning with Out-of-Sample Extrapolation60
The proposed algorithm, namely Local and Global Regressive Mapping (LGRM), employs local regression models to grasp the manifold structure and imposes a global regression term as regularization to learn a model for out-of-sample data extrapolation.
- 58
- 57
- Learning graph structure via graph convolutional networks57
A method that learns Graph Structure via graph Convolutional Networks (GSCN), which introduces the graph structure parameters measuring the correlation degrees of adjacent nodes, which is better to handle the graph-structured data of non-stationarity.
- DATA: Differentiable ArchiTecture Approximation56
Differentiable ArchiTecture Approximation (DATA) with an Ensemble Gumbel-Softmax (EGS) estimator is developed to automatically approximate architectures during searching and validating in a differentiable manner, guaranteeing the requisite efficiency during searching.
- Video Object Detection with Locally-Weighted Deformable Neighbors56
This paper proposes LWDN (Locally-Weighted Deformable Neighbors) for video object detection without utilizing time-consuming optical flow extraction networks and adopts brain-inspired memory mechanism to propagate and update the memory feature from keyframes to keyframes, called Memory-Guided Propagation.
- 55
- Blind image quality assessment via learnable attention-based pooling53
The core idea is to introduce a learnable pooling that can model human visual attention in a data-driven manner and is built by incorporating an attention module and allows for a joint learning of local quality and local weights.
- 51
- 51
- Image-Text Cross-Modal Retrieval via Modality-Specific Feature Learning49
Extensive cross-modal retrieval experiments are carried out on three challenging datasets that consist of image-document pairs or image-query click-through data from a search engine, and the results firmly demonstrate that the proposed model is much more effective.
- Embedding new data points for manifold learning via coordinate propagation49
This work first formulate this task as a quadratic programming, and then develops an iterative algorithm for coordinate propagation, to propagate the known coordinates to each of the new data points.
- Dense semantic embedding network for image captioning48
A Dense Semantic Embedding Network (DSEN) is proposed to densely embed the attributes with the multi-modal of image and text at each step of word generation and can modulate the feature distributions of the image, text modals and the hidden states to explicit semantic representation.
- 48
- 48
- 48
- 47
- Structure-Aware Convolutional Neural Networks45
By replacing the classical convolution in CNNs with the structure-aware convolution, Structure-Aware Convolutional Neural Networks (SACNNs) are readily established and there is strong evidence that SACnns outperform current models on various machine learning tasks, including image classification and clustering.
- 45
- Cooperation Does Matter: Exploring Multi-Order Bilateral Relations for Audio-Visual Segmentation43
An innovative audio-visual transformer framework, termed COMBO, an acronym for COoperation of Multi-order Bi-lateral relatiOns, which ex-plores three types of bilateral entanglements within AVS: pixel entanglement, modality entanglement, and temporal entanglement.
- 42
- 42
- 41
- 40
- 40
- 39
- Image Deblurring with Coupled Dictionary Learning37
This paper presents a learning-based framework where the knowledge hidden in huge amounts of available data is explored and exploited for image deblurring, developed under the conceptual framework of coupled dictionary learning.
- 37
- CAN: Revisiting Feature Co-Action for Click-Through Rate Prediction.36
The proposed feature Co-Action Network (CAN) can efficiently and effectively capture the feature co-action, which improves the model performance while reduce the storage and computation consumption and outperforms state-of-the-art CTR models by a large margin.
- Change detection based on auto-encoder model for VHR images35
A novel unsupervised change detection algorithm is proposed based on deep learning, where the complex correspondence between the images is established by Auto-encoder Model, by taking advantages of the powerful ability of deep learning in compensating the impacts implicitly.
- Learnable Gated Convolutional Neural Network for Semantic Segmentation in Remote-Sensing Images34
The proposed Learnable-Gated CNN (L-GCNN) showed potential for object segmentation from the RS images of buildings, impervious surfaces, low vegetation, trees and cars in urban settings, which largely varies in size and have confusing appearances.
- Cross-Modal Similarity Learning: A Low Rank Bilinear Formulation34
The proposed cross-modal similarity learning algorithm takes a bilinear formulation, and with the nuclear-norm penalization, it achieves low-rank representation and the accelerated proximal gradient algorithm is successfully imported to find the optimal solution with a fast convergence rate O(1/t2).
- 3D PostureNet: A unified framework for skeleton-based posture recognition33
An end-to-end framework based on 3D CNN, called 3D PostureNet, is developed for robust posture recognition, and achieves significantly superior performance on both skeleton-based human posture and hand posture recognition tasks.
- Geometric rectification of document images using adversarial gated unwarping network33
This model can rectify arbitrarily distorted document images with complicated page layouts and cluttered backgrounds and outperforms the state-of-the-art methods in terms of OCR accuracy and several widely used quantitative evaluation metrics.
- 33
- Semi-supervised discriminative classification with application to tumorous tissues segmentation of MR brain images33
A discriminative classification algorithm for semi-automated segmentation of brain tumorous tissues using non-parametric Bayesian Gaussian random field in the semi-supervised mode is implemented.
- Exploiting Vector Fields for Geometric Rectification of Distorted Document Images32
A segment-free method for geometric rectification of a distorted document image captured by a hand-held camera that does not have to extract curved text lines or textual blocks and can be freely applied to document images with extremely complicated page layouts and severe image quality degradation.
- HAN: An efficient hierarchical self-attention network for skeleton-based gesture recognition31
This work proposes an efficient hierarchical self-attention network (HAN) for skeleton-based gesture recognition, which is based on pure self-Attention without any CNN, RNN or GCN operators.
- Mask2Former with Improved Query for Semantic Segmentation in Remote-Sensing Images30
A Mask2Former with an improved query (IQ2Former) is proposed to enhance the capability of the query of Mask2Former by exploiting the characteristics of RS images well and overall accuracy, ablation experiment, and visualization segmentation results all indicate IQ2Former validity.
- 30
- 29
- Joint spatial-temporal attention for action recognition29
To extract robust motion representations of videos, a new spatial attention module based on 3D convolution is proposed, which can pay attention to the salient parts of the spatial areas, and a new bidirectional LSTM based temporal attention module is introduced.
- 29
- 29
- 28
- Efficient sea–land segmentation using seeds learning and edge directed graph cut28
A supervised method combining seeds learning and graph cut (GC) is proposed to address the problem of sea-land segmentation, which outperforms that of state-of-the-art methods in terms of quantitative and visual performance.
- 28
- A set theoretical method for video synopsis28
An important problem in video synopsis, that of object start-time programming, is addressed in terms of set theory, and an objective is derived to maximize visual information inVideo synopsis.
- 27
- 27
- 27
- Multi-task prompt tuning with soft context sharing for vision–language models26
Inspired by the fact that modeling task relationship by multi-task learning can usually boost performance, this work proposes a novel method SoftCPT (Soft Context Sharing for Prompt Tuning) to tune pre-trained vision-language models on multiple target few-shot tasks jointly.
- SpatioTemporal Inference Network for Precipitation Nowcasting With Multimodal Fusion25
A spatiotemporal inference network (STIN) to produce precipitation prediction from multimodal meteorological data with spatiotemporal specific filters is proposed, fusing the multimodal information, including temperature, humidity, and wind.
- Domain adaptive object detection with model-agnostic knowledge transferring25
The Knowledge Transfer Network (KTNet) is proposed, which consists of object intrinsic knowledge mining and category relational knowledge constraint modules that guide the detector to learn object-related and domain-invariant representations, and is scalable to more complex backbone networks and different detection architectures.
- Learning from the Target: Dual Prototype Network for Few Shot Semantic Segmentation25
A dual prototype network (DPNet) is proposed to dispose of few shot semantic segmentation from a new perspective and exceeds previous state-of-the-arts with a sizable margin, verifying the effectiveness of the proposed method.
- 25
- Aircraft Detection by Deep Convolutional Neural Networks25
Two types of deep learning methods are implemented, deep convolutional neural network (DNN) and deep belief net (DBN), comparing their performances with that of the traditional methods in the task of aircraft detection, and both reduce at least 40% of the false alarm rate of the traditional methods.
- No-Reference Image Quality Assessment with Reinforcement Recursive List-Wise Ranking24
This paper forms the NR-IQA as a recursive list-wise ranking problem which aims to optimize the whole quality ordering directly and improves the state-of-the-art by over 2% on the CSIQ benchmark and outperforms most compared opinion-aware models on TID2013.
- 24
- 23
- 22
- 22
- 22
- 21
- 20
- Multifocus image fusion via focus segmentation and region reconstruction20
This paper proposes three intuitive constraints on the fusion process and model them into three energy terms, i.e., reconstruction error, out-of-focus energy and smoothness regularization, which are formulated into an optimization framework problem to solve a segmentation map.
- 20
- R-4B: Incentivizing General-Purpose Auto-Thinking Capability in MLLMs via Bi-Mode Annealing and Reinforce Learning19
The central idea of R-4B is to empower the model with both thinking and non-thinking capabilities using bi-mode annealing, and apply Bi-mode Policy Optimization (BPO) to improve the model's accuracy in determining whether to activate the thinking process.
- 19
- 19
- 19
- A label propagation method using spatial-spectral consistency for hyperspectral image classification19
Through seed selection and label propagation on the graph, the problem of salt-and-pepper appearance is solved elegantly – the noisy labels are highly suppressed and most of the structures are preserved.
- Kernel sparse representation with pixel-level and region-level local feature kernels for face recognition19
This paper presents a kernel coordinate descent algorithm for the LASSO problem in the kernel space, and successfully integrates it in the SRC framework (called KCD-SRC) for face recognition, making it discriminative and robust against illumination variations and occlusions.
- 18
- 17
- 17
- VGLS: A virtual geophysical laboratory system based on C# and viustools and its application for geophysical education17
A virtual geophysical laboratory system (VGLS) based on C#, Viustools, and database technology which can make geophysical education more efficient and effective and enhance the efficiency of teaching and learning is presented.
- 17
- Contour graph based human tracking and action sequence recognition17
A new framework for human contour tracking and action sequence recognition is introduced, given a gallery of labeled human contours, where each contour is defined as a ''word'' and encoded into a contour dictionary, which is used to translate the video.
- 16
- 16
- 16
- 16
- Enhancing Visual Continual Learning with Language-Guided Supervision15
This paper revisit the role of the classifier head within the CL paradigm and replace the classifier with semantic knowledge from pretrained language models (PLMs) to generate semantic targets for each class, which are frozen and serve as supervision signals during training.
- 15
- 15
- 15
- Adversarial Gradient Driven Exploration for Deep Click-Through Rate Prediction15
A Pseudo-Exploration module is designed that simulates the model updating process after a certain item is explored and the corresponding feedback is received, and it is shown that such a process is equivalent to adding an adversarial perturbation to the model input.
- 15
- 15
- 15
- Deep generative video prediction14
A new deep generative video prediction network under the framework of generative adversarial nets is proposed and a novel cross channel color gradient (3CG) loss is designed as a guidance of deblurring.
- Pseudo low rank video representation14
A new Video Low Rank Representation model is first proposed to recover the inherent robust component of a given video, and then the recovered term is introduced to a convolutional Network (denoted pLRN) as an auxiliary pseudo Low Rank guidance.
- Deep unsupervised learning with consistent inference of latent representations14
Experimental results show that the proposed end-to-end approach can significantly improve the performance on various tasks, including image classification and clustering and indicate that the method can guide DNNs to learn more invariant feature representations in comparison with traditional unsupervised methods.
- 14
- Image deblurring with matrix regression and gradient evolution14
This paper presents a supervised learning algorithm for image deblurring that learns a matrix mapping to map each of its image patches directly to be a new one with more sharp details to be deblurred in way of gradient evolution.
- 13
- 12
- The Hitchhiker’s Guide to Autonomous Research: A Survey of Scientific Agents11
This survey bridges the existing gap by presenting a comprehensive blueprint for scientific agents’ design and introduces a unified taxonomy based on capability envelope and capability maturity, characterizing both the scope of scientific workflow coverage and the reliability of agent behavior under realistic research conditions.
- GFFNet: Global Feature Fusion Network for Semantic Segmentation of Large-Scale Remote Sensing Images11
GFFNet serves as a versatile plugin for existing RSI semantic segmentation models, particularly beneficial when the target dataset involves cropping, and proposes a cross feature fusion module that seamlessly integrates local features obtained from the convolutional network with the global contextual features.
- Dynamic High-Resolution Network for Semantic Segmentation in Remote-Sensing Images11
A Dynamic High-Resolution Network (DyHRNet), which takes HRNet as a super-architecture, aiming to leverage the important connections and channels by further investigating the parallel streams at different resolution representations of the original HRNet, and fully realizes the dynamic advantages of data adaptability.
- 11
- Spline Embedding for Nonlinear Dimensionality Reduction11
A new algorithm for nonlinear dimensionality reduction (NLDR) where smoothing splines are used to map the locally-coordinatized data points into a single global coordinate system of lower dimensionality by minimizing the low-dimensional coordinate reconstruction error.
- Defying Imbalanced Forgetting in Class Incremental Learning10
CLass-Aware Disentanglement (CLAD) is proposed as a means to predict the old classes that are more likely to be forgotten and enhance their accuracy, and can be seamlessly integrated into existing CIL methods.
- Draw an Audio: Leveraging Multi-Instruction for Video-to-Audio Synthesis10
A controllable video-to-audio synthesis model, termed Draw an Audio, which supports multiple input instructions through drawn masks and loudness signals is constructed, and extensive experiments on challenging benchmarks across two large-scale V2A datasets verify Draw an Audio achieves the state-of-the-art.
- 10
- 10
- 10
- 10
- Nonlinear dimensionality reduction with relative distance comparison10
A new algorithm for nonlinear dimensionality reduction by exploiting the local geometry of the manifold with relative distance comparisons and relaxing it to be a problem of semi-definite programming (SDP), from which a globally optimal embedding is obtained.
- 10
- UNIP: Rethinking Pre-trained Attention Patterns for Infrared Semantic Segmentation9
UNIP, a UNified Infrared Pre-training framework, is proposed, which significantly surpasses state-of-the-art (SOTA) infrared or RGB segmentation methods and demonstrates broad potential for application in other modalities, such as RGB and depth.
- TREPH: A Plug-In Topological Layer for Graph Neural Networks9
A plug-in topological layer for GNNs, termed Topological Representation with Extended Persistent Homology (TREPH), taking advantage of the uniformity of EPH, a novel aggregation mechanism is designed to collate topological features of different dimensions to the local positions determining their living processes.
- 9
- 9
- 9
- IF-Bench: Benchmarking and Enhancing MLLMs for Infrared Images with Generative Visual Prompting8
This work introduces IF-Bench, the first high-quality benchmark designed for evaluating multimodal understanding of infrared images and proposes a training-free generative visual prompting (GenViP) method, which leverages advanced image editing models to translate infrared images into semantically and spatially aligned RGB counterparts, thereby mitigating domain distribution shifts.
- 8
- Rethinking Comprehensive Benchmark for Chart Understanding: A Perspective from Scientific Literature8
A new benchmark, Scientific Chart QA (SCI-CQA), is introduced, which emphasizes flowcharts as a critical yet often overlooked category, and context-based chart understanding is explored, highlighting the crucial role of contextual information in solving previously unanswerable questions.
- 8
- 8
- SafeNet: Scale-normalization and Anchor-based Feature Extraction Network for Person Re-identification8
The advantage of jointly utilizing multi-scale abstract information to learn powerful features over full body and parts is shown, and an anchor-based method to capture the local contents by stacking convolutions of kernels with various aspect ratios is proposed.
- Large Scale Image Annotation via Deep Representation Learning and Tag Embedding Learning8
A novel model based on deep representation learning and tag embedding learning is proposed and an objective function modeling triplet relationships (irrelevant tag, image, relevant tag) is proposed with maximum margin pursuit.
- 8
- 8
- Graph Aggregating-Repelling Network: Do Not Trust All Neighbors in Heterophilic Graphs7
A novel framework named GARN is proposed to effectively extract both homophilic and heterophilic information from graph spectral and spatial theory and indicates the effectiveness of the proposed framework and mechanism over several representative GNN baselines.
- 7
- 7
- Combining Discrete and Continuous Representation: Scale-Arbitrary Super-Resolution for Satellite Images7
A novel image representation that combines the discrete and continuous representation, known as CDCR, which enables the extension of images to any desired resolution in a plug-and-play manner and introduces a scale cumulative ascent (SCA) method.
- PAD: Self-Supervised Pre-Training with Patchwise-Scale Adapter for Infrared Images7
A pre-training paradigm called Pre-training with ADapter (PAD), which uses adapters to learn domain-specific features while freezing parameters pre-trained on ImageNet to retain the general feature extraction capability is proposed, applicable to any transformer-based SSL method.
- 7
- Do We Need Binary Features for 3D Reconstruction?7
It is revealed that not all binary features are capable of 3D reconstruction and most of them are inferior to the classical SIFT based method in terms of reconstruction accuracy and completeness with a not significant better computational performance.
- Active Rectification of Curved Document Images Using Structured Beams7
This method uses structured beams projected upon the curved document page to recover the spatial curves on it and uses a Lambertian illumination model as guidance to rectify the non-uniform illumination in the images.
- 7
- Subspace Regularization: A New Semi-supervised Learning Method7
This work introduces into semi-supervised learning the classic low-dimensionality embedding assumption, stating that most geometric information of high dimensional data is embedded in a low dimensional manifold.
- Dynamic modulation transfer function measurement of image intensifiers using a narrow slit7
The authors' numerical results show that with the increase of the reverse bias of MCPIIs, the resolution is improved rapidly below 40 V and then gradually decreased, while with the MCP bias increased, both M CPIIs start to suffer rapid reductions in resolution at 800 and 750 V.
- Contour Matching Based on Belief Propagation7
This paper uses graphical model based probabilistic inference methods to solve the problem of contour matching, which is a fundamental problem in computer vision and comparable experimental results show that better correspondences can be achieved.
- Re-ranking Reasoning Context with Tree Search Makes Large Vision-Language Models Stronger6
A multimodal RAG framework, termed RCTS, is proposed, which enhances LVLMs by constructing a Reasoning Context-enriched knowledge base and a Tree Search re-ranking method, and proposes a Monte Carlo Tree Search with Heuristic Rewards (MCTS-HR) to prioritize the most relevant examples.
- Siamese InternImage for Change Detection6
By introducing deformable convolution into Siamese InternImage architecture, the proposed CNN-based change detection vision foundation model is capable of capturing long-range dependencies and global information.
- AVESFormer: Efficient Transformer Design for Real-Time Audio-Visual Segmentation6
This paper introduces AVESFormer, the first real-time Audio-Visual Efficient Segmentation transformer that achieves fast, efficient and light-weight simultaneously, and proposes ELF decoder to bring greater efficiency by facilitating convolutions suitable for local features to reduce computational burdens.
- Reusable Architecture Growth for Continual Stereo Matching6
This work proposes to perform continual stereo matching where a model is tasked to continually learn new scenes, overcome forgetting previously learned scenes, and continuously predict disparities at inference, by introducing a Reusable Architecture Growth (RAG) framework.
- 6
- 6
- 6
- Image tag-ranking via pairwise supervision based semi-supervised model6
A novel model which uses images with ranked tag lists as its supervision information and sufficiently exploits the tag relevance to images as well as the ranking structures of tag lists to capture the intrinsic ranking structures.
- 6
- Texture Image Segmentation: An Interactive Framework Based on Adaptive Features and Transductive Learning6
An interactive framework for texture segmentation that the user can define the textures to be segmented by labelling a small part of points belonging to them and further improve the segmentation quality through a few interactive manipulations if necessary.
- 5
- 5
- Graph convolutional network with tree-guided anisotropic message passing5
This paper proposes a tree-guided anisotropic GCN, which applies an anisotrop aggregation strategy with competitive expressiveness and a large receptive field, and outperforms baseline methods and recent works on several synthetic benchmarks and datasets from different real-world tasks.
- 5
- 5
- 5
- 3D Point Cloud Classification Based on Discrete Conditional Random Field5
This paper uses the local geometric shape features to construct the CRF, including the nearest neighbor tetrahedral volume, Gaussian curvature, the neighbourhood normal vector consistency and the neighbourhood minimum principal curvature direction consistency, and proposes four discrete criteria for CRF parameter optimization.
- Co-Training Based Segmentation of Merged Moving Objects5
A novel method for segmenting merged objects into individual ones, based on a naive Bayes classifier and an appearance classifier based on Multi-block Local Binary Pattern features is proposed.
- MeteorPred: A Meteorological Multimodal Large Model and Dataset for Severe Weather Event Prediction4
The first large-scale multimodal dataset for severe weather events prediction, comprising 421,363 pairs of raw multi-year meteorological data and corresponding text caption, and a Meteorology Multimodal Large Model that directly ingests 4D meteorological inputs are introduced.
- 4
- Efficient redundancy reduction for open-vocabulary semantic segmentation4
ERR-Seg is proposed, an efficient architecture that incorporates Redundancy-Reduced Hierarchical Cost maps (RRHC) and Redundancy-Reduced Cost Aggregation (RRCA) and results in a lightweight structure for OVSS, characterized by substantial memory and computational savings without compromising accuracy.
- Multi-level consistency regularization for domain adaptive object detection4
This work explores the possibility of transferring detectors with only source domain data and style information of the target domain with favorable performance compared with existing state-of-the-art methods, and proposes three consistency regularizations to enhance the adaptation performance of the detector.
- 4
- Subgraph-aware graph structure revision for spatial–temporal graph modeling4
A subgraph-aware structure revision graph convolution module (SASR-GCM) is designed, which revises the learned stable graph to obtain a dynamic one to automatically infer the dynamics of spatial correlations and can efficiently capture the spatial-temporal correlations and extract complex spatial- Temporal graph features.
- 4
- 4
- Parameter optimization criteria guided 3D point cloud classification4
Experimental results show that the proposed three discrete criteria for parameter optimization to design explicit functions can be applied to the classification of the 3D point cloud of the scene, and can be used to improve the classification accuracy of small-scale point sets when different classes have great disparities in the number.
- 4
- 4
- 4
- Image Guided Tone Mapping with Locally Nonlinear Model4
An effective locally nonlinear tone mapping algorithm for compressing the High Dynamic Range (HDR) images by introducing local gamma correction with adaptive parameters on small overlapping patches over the entire input image.
- 4
- Interactive Visual Object Extraction Based on Belief Propagation4
An interactive framework which can allow the user to label multiply foreground objects of interest and some applications in color transfer, image completion and motion detection are given in this paper.
- A New Approach to Human Motion Sequence Recognition with Application to Diving Actions4
This paper proposes an approach to human motion sequence recognition based on 2D spatio-temporal shape analysis, which is used to identify diving actions and shows that high correct recognition ratio can be achieved.
- SAM-MI: A Mask-injected Framework for Enhancing Open-vocabulary Semantic Segmentation with SAM3
A novel mask-injected framework, SAM-MI, which effectively integrates SAM with OVSS models to address SAM’s tendency to over-segment, and introduces shallow mask aggregation (SMAgg) to merge partial masks to mitigate SAM’s over-segmentation issue.
- 3
- 3
- Compositional Kronecker Context Optimization for vision–language models3
Compositional Kronecker Context Optimization achieves state-of-the-art performance under base-to-new, domain and cross-task generalization evaluation, but also has the metrics of fewer learnable parameters and efficient training and inference speed.
- Few-shot video object segmentation with prototype evolution3
Experimental results on the Youtube-VIS dataset of 2019 version and 2021 version demonstrate that the proposed PENet outperforms the previous methods with a sizable margin, validating the superiority of the proposed model.
- Calibrated Cache Model for Few-Shot Vision-Language Model Adaptation3
Three calibration modules aimed at addressing the gap between pre-training and adaptation of vision-language models and the high complexity of GPs are presented and both training-free and training-required variants are proposed.
- 3
- Exploring Universal Principles for Graph Contrastive Learning: A Statistical Perspective3
The comparison experiments with current state-of-the-arts demonstrate the effectiveness and sufficiency of two principles for high-quality graph representations and constructs a general self-supervised framework for negative-free graph contrastive learning.
- 3
- 3
- Ltaf-Net: Learning Task-Aware Adaptive Features and Refining Mask for Few-Shot Semantic Segmentation3
- 3
- 3
- Incremental Poisson Surface Reconstruction for Large Scale Three-Dimensional Modeling3
A novel Incremental Poisson Surface Reconstruction method based on point clouds and the adaptive octree and providing an octree node classification mechanism to achieve the seamless reconstruction between adjacent blocks is proposed.
- 3
- 3
- WS-R-IR Adapter: A Multimodal RGB–Infrared Remote Sensing Framework for Water Surface Object Detection2
The WS-R-IR Adapter is proposed, a parameter-efficient vision foundation model (VFM)-based framework for shipborne RGB–IR object detection that adapts frozen VFM representations via lightweight task-specific designs and demonstrates the effectiveness of parameter-efficient VFM adaptation for multimodal water surface remote sensing.
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- What and Where the Themes Dominate in Image2
A novel framework is proposed, which explicitly applies the known salient objects in image captioning to address the exposure bias problem between the training and the testing modes and formulates two constraints in the theme learning process to encourage distinctiveness between different images.
- 2
- Facade Labeling via Explicit Matrix Factorization2
An alternating optimization algorithm is thus developed to solve the matrix factorization problem, where dynamic programming is used to optimize the block matrices.
- 2
- Cross Modal Deep Model and Gaussian Process Based Model for MSR-Bing Challenge2
A regression based cross modal deep learning model and a Gaussian Process scoring model are proposed that can score the query-image pairs based on the relevance between queries and images in the MSR-Bing Image Retrieval Challenge.
- 2
- 2
- Soft Constraint Harmonic Energy Minimization for Transductive Learning and its Two Interpretations2
This paper extends the harmonic energy minimization algorithm and proposes a novel transductive learning algorithm on graph with soft label and soft constraint, which can be interpreted from the view of label propagation and a special random walks on graph, which make the algorithm intuitively reasonable.
- 2
- PA-Net: Precipitation-Adaptive Mixture-of-Experts for Long-Tail Rainfall Nowcasting1
The Precipitation-Adaptive Network (PA-Net) is proposed, a Transformer framework whose computational budget is explicitly governed by rainfall intensity, and demonstrates consistent improvements over state-of-the-art baselines, with particularly significant gains in heavy-rain and rainstorm regimes.
- Refer-ASV: Referring Multi-Object Tracking in Autonomous Surface Vehicle Navigation Scenes1
Refer-ASV is introduced, the first RMOT dataset tailored for ASV navigation in complex water-surface scenes and RAMOT is proposed, an end-to-end baseline framework that enhances visual–language alignment throughout the tracking pipeline by improving visual–language alignment and robustness in challenging maritime environments.
- 1
- 1
- A Survey of Low-shot Vision-Language Model Adaptation via Representer Theorem1
A unified computational framework is proposed from the perspective of Representer Theorem and then derives many of the existing methods by specializing this framework by exploiting the closed-form solution of kernel ridge regression.
- STACnovGRU: weather forecasting based on spatio-temporal adaptive convolutional GRU1
A Spatio-Temporal Adaptive Convolution for the Gated Recurrent Unit (GRU) to improve the ability of extracting spatio-temporal features and develop additional convolution layers under the framework of the ordinary RNN to learn simultaneously the sampling positions and weights of convolutional kernels.
- 1
- Robust Feature Rectification of Pretrained Vision Models for Object Recognition1
ROFER first estimates the type and intensity of the degradation that corrupts the image features, then leverages a Fully Convolutional Network (FCN) to rectify the features from the degradation by pulling them back to clear features.
- 1
- 1
- 1
- 1
- Learning Motion Priors for Efficient Video Object Detection.1
This paper first initialize some motion priors for each location and then use them to propagate features across frames and achieves state-of-the-art performance on the ImageNet VID dataset with real-time speed.
- 1
- Fine-structured object segmentation via neighborhood propagation1
A novel method for fine-structured object segmentation via label propagation, which introduces a novel neighborhood system combining both local and nonlocal connections, with a robust scheme for edge weight calculation, is presented.
- 1
- 1
- 1
- 1
- 1
- Transferring Segmentation from Image to Image via Contextual Sparse Representation1
This work proposes a transferring segmentation framework, which aims to automatically segment new images when a single segmented example is given, and proposes to construct a contextual dictionary by incorporating neighboring information as context.
- The application of auto-gated power supply in ICCD camera1
The auto-gated power supply method put forward to implementing automatic brightness control (ABC) helps to improve the image quality of the ICCD camera and improves the dynamic range of the camera.
- 1
- 1
- 1
- 1
- 1
- 1
- –
- –
- –
- –
- MeTok: An Efficient Meteorological Tokenization with Hyper-Aligned Group Learning for Precipitation Nowcasting–
This paper focuses primarily on the task of precipitation nowcasting and develops an efficient distribution-centric Meteorological Tokenization (MeTok) scheme, which spatially sequences to group similar meteorological features, demonstrating scalability, stability, and superiority over traditional methods.
- –
- InfEngine: A Self-Verifying and Self-Optimizing Intelligent Engine for Infrared Radiation Computing–
This work introduces InfEngine, an autonomous intelligent computational engine designed to drive a paradigm shift from human-led orchestration to collaborative automation, and illustrates how researchers can transition from manual coding to collaborating with self-verifying, self-optimizing computational partners.
- Beyond Counting: Evaluating Abstract and Emotional Reasoning in Vision-Language Models–
EmojiGrid provides a quantifiable, fine-grained benchmark to diagnose VLM limitations and guides future progress toward models that can truly perceive, reason about, and interpret complex, symbol-rich visual scenes.
- –
- –
- Advancing Scientific Chart Understanding: The SCI-CQA Benchmark and Beyond–
An in-depth study on scientific chart understanding finds that fine-grained visual perception is essential for accurate chart comprehension, and proposes the Visual Insight Extraction Module (VIEM), which leverages saliency cues-such as keypoints and OCR signals-to highlight essential visual details.
- –
- Audio-Visual Segmentation via Depth-Guided Collaborative Modeling–
This study designs a Depth-Aware Dynamic Modulator to improve the separation of adjacent objects while preserving intra-object feature consistency and proposes DGCM-AVS, a tri-modal framework that jointly models audio, visual, and depth information.
- LookFlow: Training-Free and Efficient High-Resolution Image Synthesis via Dynamic Lookahead Guidance Flow–
LookFlow is a training-free high-resolution synthesis framework that accelerates inference while preserving visual quality, and employs a dynamic lookahead guidance flow mechanism to refine high-resolution velocity predictions by leveraging multi-timestep lookahead information extracted from a low-resolution flow.
- –
- A Dual-UNet Diffusion Framework for Personalized Panoramic Generation–
A diffusion-based framework for customized multi-view image generation that introduces a decoupled feature injection mechanism within a dual-UNet architecture to handle weakly correlated reference images, effectively integrating spatial information by concurrently feeding both reference images and noise into the denoising branch.
- –
- –
- Free Lunch for Generating Effective Outlier Supervision–
This paper investigates the OOD detection problem based on the Bayes rule and proposes an ultra-effective method to generate near-realistic outlier supervision, which significantly reduces the FPR95 over 12.50\% compared with the previous schemes, boosting the reliability of machine learning systems.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Guided tone mapping–
A simple but effective tone mapping operator is proposed, which can achieve impressive results with low complexity in both the formulation and computation of the operator.
- –
- –
- Semi-supervised Gaussian random field transduction and induction–
A semi-supervised inductive algorithm adopting a Gaussian random field (GRF) and Gaussian process and a new conditional probability called the extended Bernoulli model (EBM) is proposed, which generalizes the logistic regression to the semi- supervised case.
- –
- Exemplar-Based Human Contour Tracking–
An exemplar-based tracking framework for human contour tracking with main task to generate contours to fill in the gaps in-between in the test sequences, and to supply the dynamics for updating the target contour by fast contour query.
- –
- –
- –
- –
- –
- –
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 2026-10-15. 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.