Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Yulan GuoSuggested from co-authorship
- Wei AnSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works122 from public data
- Dense Nested Attention Network for Infrared Small Target Detection1,048
This paper designs a dense nested interactive module (DNIM) to achieve progressive interaction among high-level and low-level features and proposes a cascaded channel and spatial attention module (CSAM) to adaptively enhance multi- level features.
- Unsupervised Degradation Representation Learning for Blind Super-Resolution419
This paper proposes an unsupervised degradation representation learning scheme for blind SR without explicit degradation estimation, and introduces a Degradation-Aware SR (DASR) network with flexible adaption to various degradations based on the learned representations.
- 320
- Learning Parallax Attention for Stereo Image Super-Resolution320
A parallax-attention mechanism with a global receptive field along the epipolar line to handle different stereo images with large disparity variations is introduced and a new and the largest dataset for stereo image SR is proposed.
- Spatial-Angular Interaction for Light Field Image Super-Resolution208
Experimental results demonstrate the superiority of LF-InterNet over the state-of-the-art methods, i.e., the method can achieve high PSNR and SSIM scores with low computational cost, and recover faithful details in the reconstructed images.
- Light Field Image Super-Resolution With Transformers190
This paper proposes a simple but effective Transformer-based method for LF image SR, designed to incorporate complementary information among different views, and achieves superior SR performance with a small model size and low computational cost.
- Deep Video Super-Resolution Using HR Optical Flow Estimation176
This paper proposes an end-to-end video SR network to super-resolve both optical flows and images and shows that the network achieves the state-of-the-art performance.
- Deformable 3D Convolution for Video Super-Resolution170
A deformable 3D convolution network (D3Dnet) is proposed to incorporate spatio-temporal information from both spatial and temporal dimensions for video SR, and achieves state-of-the-art SR performance.
- Learning A Single Network for Scale-Arbitrary Super-Resolution162
This paper develops a plug-in module for existing SR networks to perform scale-arbitrary SR, which consists of multiple scale-aware feature adaption blocks and a scale- aware upsampling layer that enables the network to adapt to arbitrary scale factors.
- Parallax Attention for Unsupervised Stereo Correspondence Learning159
A generic parallax-attention mechanism (PAM) to capture stereo correspondence regardless of disparity variations is proposed and Experimental results show that the PAM is generic and can effectively learn stereo correspondence under large disparity variations in an unsupervised manner.
- 153
- A Stereo Attention Module for Stereo Image Super-Resolution143
Using SAM, the generic stereo attention module (SAM) is proposed to extend arbitrary SISR networks to stereo images and can exploit cross-view information while maintaining the superiority of intra-View information exploitation, resulting in notable performance gain to SisR networks.
- Occlusion-Aware Cost Constructor for Light Field Depth Estimation114
A simple and fast cost constructor to construct matching cost for LF depth estimation, composed by a series of convolutions with specifically designed dilation rates, which is occlusion-aware and can handle occlusions by dynamically modulating pixels from different views.
- Symmetric Parallax Attention for Stereo Image Super-Resolution112
This paper proposes a symmetric bi-directional parallax attention module (biPAM) and an inline occlusion handling scheme to effectively interact cross-view information and designs a Siamese network equipped with a biPAM to super-resolve both sides of views in a highly symmetric manner.
- 107
- Learning for Video Super-Resolution Through HR Optical Flow Estimation107
This paper proposes an end-to-end trainable video SR framework to super-resolve both images and optical flows and demonstrates that HR optical flows provide more accurate correspondences than their LR counterparts and improve both accuracy and consistency performance.
- Learning Non-Local Spatial-Angular Correlation for Light Field Image Super-Resolution102
This paper adopts the epipolar plane image (EPI) representation, develops a Transformer network with repetitive self-attention operations to learn the spatial-angular correlation, and develops a Transformer network with repetitive self-attention operations to learn the spatial-angular correlation.
- DeOccNet: Learning to See Through Foreground Occlusions in Light Fields94
This paper handles the LF de-occlusion (LF-DeOcc) problem using a deep encoder-decoder network (namely, DeOccNet), and is the first deep learning-based LF-De Occ method.
- 80
- A Systematic Survey of Deep Learning-Based Single-Image Super-Resolution79
This survey gives an overview of DL-based SISR methods and group them according to their design targets and presents the reconstruction results of some classic SISR methods to intuitively know their performance.
- Decoupling Makes Weakly Supervised Local Feature Better74
This paper proposes a decoupled training describe-then-detect pipeline tailored for weakly supervised local feature learning, and introduces a line-to-window search strategy to explicitly use the camera pose information for better descriptor learning.
- Motion and Appearance Decoupling Representation for Event Cameras61
Inspired by human visual prior, a motion and appearance decoupling (MAD) event representation is proposed to disentangle the mixed spatial-temporal event tensor into two independent branches to reduce the network’s learning burden toward complex high-level interpretation tasks.
- Monte Carlo Linear Clustering with Single-Point Supervision is Enough for Infrared Small Target Detection60
This work proposes the first method to achieve SIRST detection with single-point supervision, and can accurately recover pseudo masks and thus turn arbitrary fully supervised SIRST detection networks into weakly supervised ones with only single point annotation.
- Local Motion and Contrast Priors Driven Deep Network for Infrared Small Target Superresolution59
This article proposes the first infrared small target SR method named local motion and contrast prior driven deep network (MoCoPnet) to integrate the domain knowledge of therared small target into deep network, which can mitigate the intrinsic feature scarcity of infrared small targets.
- Lightweight Pixel Difference Networks for Efficient Visual Representation Learning58
Two novel types of convolutions are proposed, dubbed Pixel Difference Convolution (PDC) and Binary PDC (Bi-PDC), which enjoy the following benefits: capturing higher-order local differential information, computationally efficient, and able to be integrated with existing DNNs.
- Gated Recurrent Multiattention Network for VHR Remote Sensing Image Classification57
A gated recurrent multiattention neural network (GRMA-Net) that uses multilevel attention modules to focus on informative regions to extract more discriminative features and is fed into a deep-gated recurrent unit (GRU) to capture long-range dependency and contextual relationship.
- Real-World Light Field Image Super-Resolution Via Degradation Modulation55
Compared with existing state-of-the-art single and LF image SR methods, the proposed method achieves superior SR performance under a wide range of degradation, and generalizes better to real LF images.
- 55
- NTIRE 2023 Challenge on Light Field Image Super-Resolution: Dataset, Methods and Results52
The first NTIRE challenge on light field (LF) image super-resolution (SR), which aims at super-resolving LF images under the standard bicubic degradation with a magnification factor of 4, develops a new LF dataset called NTIRE-2023 for validation and test, and provides a toolbox called BasicLFSR to facilitate model development.
- 50
- Deep Bilateral Learning for Stereo Image Super-Resolution50
This letter uses a parallax-attention module to incorporate information from left and right views to learn content-aware bilateral filters, which are used to recover missing details at different spatial locations while preserving stereo consistency in a bilateral stereo super-resolution network.
- VideoDirector: Precise Video Editing via Text-to-Video Models49
Experimental results demonstrate that the method effectively harnesses the powerful temporal generation capabilities of T2V models, producing edited videos with state-of-the-art performance in accuracy, motion smoothness, realism, and fidelity to unedited content.
- 48
- IF-ConvTransformer46
An HAR framework composed of an Inertial Measurement Unit (IMU) fusion block and an applied ConvTransformer subnet is proposed in this paper, Inspired by the complementary filter, the IMU fusion block performs multi-modal fusion of commonly used sensors according to their physical relationships.
- 44
- DropoutGS: Dropping Out Gaussians for Better Sparse-view Rendering40
A Random Dropout Regularization (RDR) is proposed to exploit the advantages of low-complexity models to alleviate overfitting and an Edge-guided Splitting Strategy (ESS) is developed to remedy the lack of high-frequency details for these models.
- 39
- 39
- 38
- Heterogeneous Graph Transformer for Multiple Tiny Object Tracking in RGB-T Videos38
This paper proposes a novel framework called HGT-Track (Heterogeneous Graph Transformer based Multi-Tiny-Object Tracking), and introduces the first benchmark VT-Tiny-MOT (Visible-Thermal Tiny Multi-Object Tracking) for RGB-T fused multiple tiny object tracking.
- PointCMP: Contrastive Mask Prediction for Self-supervised Learning on Point Cloud Videos37
The proposed contrastive mask prediction (PointCMP) framework employs a two-branch structure to achieve simultaneous learning of both local and global spatiotemporal information and generates hard samples to facilitate learning representations with better discrimi-nation and generalization performance.
- NTIRE 2022 Challenge on Stereo Image Super-Resolution: Methods and Results35
The 1st NTIRE challenge on stereo image super-resolution (restoration of rich details in a pair of low-resolution stereo images) with a focus on new solutions and results establishes a new benchmark for stereo image SR.
- SpecDETR: A transformer-based hyperspectral point object detection network33
This paper introduces SpecDETR, the first specialized network for hyperspectral multi-class point object detection, which eliminates dependence on pre-trained backbone networks commonly required by vision-based object detectors and outperforms SOTA visual object detection networks and HTD methods.
- Masked Spatio-Temporal Structure Prediction for Self-supervised Learning on Point Cloud Videos33
MaST-Pre is a Masked Spatio-Temporal Structure Prediction method to capture the structure of point cloud videos without human annotations and consists of two self-supervised learning tasks that are forced to model the spatial and temporal structure in point cloud videos.
- Probing Deep Into Temporal Profile Makes the Infrared Small Target Detector Much Better29
This paper remodels the IRST detection task as a one-dimensional signal anomaly detection task, and proposes an efficient deep temporal probe network (DeepPro) that only performs calculations in the time dimension for IRST detection, which outperforms existing state-of-the-art IRST detection methods on widely-used benchmarks with extremely high efficiency.
- 28
- Point Contrastive Prediction with Semantic Clustering for Self-Supervised Learning on Point Cloud Videos27
This paper proposes a selection strategy to retain proper negatives and make use of high-similarity samples from other instances as positive supplements for contrastive learning at the point level, and introduces a new pretext task by achieving semantic alignment of superpoints.
- 3D Face Recognition: Two Decades of Progress and Prospects26
A comprehensive review of the progress achieved by both traditional and deep learning–based 3D face recognition methods in the last two decades is presented and comparative results on several publicly available datasets under different challenges of facial expressions, pose variations, and occlusions are presented.
- SaMam: Style-aware State Space Model for Arbitrary Image Style Transfer25
A Mamba-based style transfer framework that outperforms state-of-the-art methods in terms of both accuracy and efficiency, and introduces local enhancement and zigzag scan mechanisms to address the problems of local pixel forgetting, channel redundancy and spatial discontinuity of existing SSMs.
- 25
- 24
- Deep Semantic Graph Matching for Large-Scale Outdoor Point Cloud Registration24
This article treats the point cloud registration problem as a semantic instance matching and registration task and proposes a deep semantic graph matching method (DeepSGM) for large-scale outdoor point cloud registration that improves the registration performance and outperforms various state-of-the-art methods.
- 24
- Learning Robust Stereo Matching in the Wild with Selective Mixture-of-Experts23
This work proposes SMoEStereo, a novel framework that adapts VFMs for stereo matching through a tailored, scene-specific fusion of Low-Rank Adaptation and Mixture-of-Experts (MoE) modules, and proposes a lightweight decision network that selectively activates MoE modules based on input complexity, balancing efficiency with accuracy.
- Unsupervised Degradation Representation Learning for Unpaired Restoration of Images and Point Clouds23
- NTIRE 2024 Challenge on Stereo Image Super-Resolution: Methods and Results23
This paper summarizes the 3rd NTIRE challenge on stereo image super-resolution with a focus on new solutions and results and establishes a new benchmark for stereo image SR.
- 21
- 19
- 18
- Layout2Scene: 3D Semantic Layout Guided Scene Generation via Geometry and Appearance Diffusion Priors17
A text-to-scene generation method using additional semantic layout as the prompt to inject precise control of 3D object positions and can generate more plausible and realistic scenes as compared to state-of-the-art approaches.
- 17
- 16
- 16
- 16
- 15
- 15
- Panoptic segmentation of 3D point clouds with Gaussian mixture model in outdoor scenes15
This work introduces a framework for modeling points belonging to the same instance using learnable Gaussian distributions and formulate the point cloud as a Gaussian mixture model and introduces a unified loss function that links the embedding learning and instance clustering in an end-to-end manner.
- 15
- 14
- 14
- 11
- 10
- 10
- Learning for Scale-Arbitrary Super-Resolution from Scale-Specific Networks9
This paper proposes a plug-in module for existing SR networks to perform scale-arbitrary SR, which consists of multiple scale- aware feature adaption blocks and a scale-aware upsampling layer, and introduces a Scale-aware knowledge transfer paradigm to transfer knowledge from scale-specific networks to the scale-Arbitrary network.
- 9
- 7
- 6
- Tangram-Splatting: Optimizing 3D Gaussian Splatting Through Tangram-inspired Shape Priors6
This work introduces a novel methodology (Tangram-Splatting) that leverages shape priors to optimize 3D scene fitting and achieves a remarkable average reduction in memory consumption used to store optimized parameters and decreases the training time by at least 10 minutes.
- 6
- Flickr1024: A Dataset for Stereo Image Super-Resolution.6
A large-scale stereo dataset named Flickr1024 is proposed that can improve the performance of stereo SR algorithms and is trained on the KITTI2015, Middlebury, and Flickr1024 datasets.
- Infrared point target detection based on multi-label generative MRF model6
The experimental results demonstrate that the detection method based on MG-MRF has a strong suppression of false alarm with superior performance in terms of accuracy and efficiency.
- Deep Lookup Network5
This paper introduces a generic and efficient lookup operation which can be used as a basic operation for the construction of neural networks and develops lookup networks for the image classification, image super-resolution, and point cloud classification tasks.
- A Unified Hierarchical Framework for Fine-grained Cross-view Geo-localization over Large-scale Scenarios5
This paper proposes UnifyGeo, a novel unified hierarchical geo-localization framework that integrates retrieval and metric localization tasks into a single network, and designs a re-ranking mechanism guided by a dedicated loss function, which enhances geo-localization performance by improving both retrieval accuracy and metric localization references.
- Learning Sparse Masks for Efficient Image Super-Resolution5
An SR network (SMSR) is developed to learn sparse masks to prune redundant computation conditioned on the input image so that redundant computation can be accurately located and skipped while maintaining comparable performance.
- 4
- 4
- 4
- 4
- 4
- PanMatch: Unleashing the Potential of Large Vision Models for Unified Matching Models3
This work presents PanMatch, a versatile foundation model for robust correspondence matching that achieves multi-task integration by endowing displacement estimation algorithms with unprecedented generalization capabilities and presents unprecedented zero-shot performance in abnormal scenarios, such as rainy day and satellite imagery.
- 3
- 3
- Point target detection utilizing super-resolution strategy for infrared scanning oversampling system3
A point target detection method utilizing super-resolution (SR) strategy concerning infrared scanning oversampling system is formulates, with an accelerated SR strategy proposed to realize fast de-aliasing of the oversampled image and an adaptive MRF-based regularization designed to achieve the preserving and aggregation of target energy.
- 2
- 2
- 2
- Adaptive Spectral Denoising Network for Efficient Maneuvering Target Tracking2
This work proposes the adaptive spectral denoising network (ASDN), comprising an adaptive spectral denoising filter (ASDF) and a temporal local-pattern guided block (TLGB), and introduces large-scale maneuvering target trajectory (LASTv2), a large-scale dataset of maneuvering target trajectories characterized by diverse and complex motion patterns, specifically designed to simulate real-world adversarial scenarios with high mobility and stochastic dynamics.
- Pluggable Style Representation Learning for Multi-style Transfer2
A style-aware multi-style transfer network (SaMST) to adapt to diverse styles using pluggable style representations and is able to accommodate diverse image styles in the learned style representations without introducing additional overhead during inference, thereby maintaining efficiency.
- ARBiBench: Benchmarking Adversarial Robustness of Binarized Neural Networks2
ARBiBench, a comprehensive benchmark to evaluate the robustness of BNNs against adversarial perturbations on CIFAR-10 and ImageNet, reveals that the adversarial robustness of BNNs exhibits a completely opposite performance on the two datasets under white-box attacks.
- 1
- 1
- 1
- 1
- 1
- Event‐based motion segmentation of small objects in the wild1
An event-based motion segmentation method to segment moving small objects in events obtained from the wild, namely EMSS is proposed and the first dataset for event-based motion segmentation of small objects is developed, namely EMSS.
- 1
- An infrared small target detection method based on visual spatio-temporal local contrast measure1
Experimental results demonstrate that the proposed infrared small target detection method based on visual spatio-temporal contrast measure (STLCM) is superior in compressing the false alarm and efficient in different conditions.
- –
- –
- Towards UAV Detection in the Real World: A New Multispectral Dataset UAVNet-MS and a New Method–
This work introduces UAVNet-MS, the first multispectral dataset for fine-grained small-UAV detection, and proposes MFDNet, a dual-stream baseline addressing array-induced parallax and spatial-spectral fusion that achieves +6.2% AP50 improvement over best RGB-only methods.
- –
- Triple Spectral Fusion for Sensor-Based Human Activity Recognition–
This paper proposes a novel triple spectral fusion framework for sensor-based human activity recognition that uses adaptive filtering in the Fourier, graph Fourier, and wavelet domains, enabling effective multi-sensor fusion and context correlation.
- UPGS: Unified Pose-aware Gaussian Splatting for Dynamic Scene Deblurring–
This work recast camera and object motion as per-primitive SE(3) affine transformations on 3D Gaussians and formulate a unified optimization objective that achieves significant gains in reconstruction quality and pose estimation accuracy over prior dynamic deblurring methods.
- –
- Traffic Image Restoration under Adverse Weather via Frequency-Aware Mamba–
Frequency-Aware Mamba is proposed, a novel framework that integrates frequency guidance with sequence modeling for efficient image restoration and a novel Adaptive Frequency Scanning Mechanism for the Mamba, which enables the Mamba to achieve frequency-domain scanning across distinct subgraphs, thereby fully leveraging the texture distribution characteristics inherent in subgraph structures.
- –
- Preserving Full Degradation Details for Blind Image Super-Resolution–
This paper introduces a degradation-aware Mamba module to efficiently model long-range dependency between the anchor pixel and the surrounding informative pixels and develops a distribution alignment loss to facilitate the learning of the degradation representations by introducing a bounded constraint.
- –
- –
- –
- –
- The optimization design and robust test of TVF for moving target detection–
The simulation results show that the relative high signal-to-clutter gain can be obtained in different regions, which satisfies the requirement of target detection algorithm in high frame rate detection system.
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.