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
- Weisi LinSuggested from co-authorship
- Chia‐Wen LinSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works376 from public data
- 647
- 496
- 362
- Dense Attention Fluid Network for Salient Object Detection in Optical Remote Sensing Images355
An end-to-end Dense Attention Fluid Network (DAFNet) for SOD in optical RSIs that significantly outperforms the existing state-of-the-art SOD competitors is proposed and a new and challenging optical RSI dataset is constructed that contains 2,000 images with pixel-wise saliency annotations.
- 334
- Single Image Deraining: From Model-Based to Data-Driven and Beyond289
This paper presents milestones of single-image deraining methods, reviews a broad selection of previous works in different categories, and provides insights on the historical development route from the model-based to data-driven methods.
- 242
- 239
- 238
- 235
- 231
- 225
- 215
- 208
- Weakly Supervised Video Anomaly Detection via Center-Guided Discriminative Learning197
An anomaly detection framework, called Anomaly Regression Net (ARNet), which only requires video-level labels in training stage is proposed, which yields a new state-of-the-art result for video anomaly detection on ShanghaiTech dataset.
- 195
- 174
- 166
- No-Reference and Robust Image Sharpness Evaluation Based on Multiscale Spatial and Spectral Features146
- A visually secure image encryption scheme based on semi-tensor product compressed sensing123
The simulations show that the proposed visual security strategy based on STP-CS can not only greatly reduce the storage space occupied by the measurement matrix but also maintain high reconstruction quality compared with the traditional CS model while ensuring visual security quality.
- 123
- 123
- Abnormal event detection in crowded scenes based on deep learning108
Experimental results show that the proposed abnormal event detection method can obtain much better performance than the existing ones on the public video database.
- High-order local ternary patterns with locality preserving projection for smoke detection and image classification101
This work proposes a novel feature extraction method by encoding high order directional derivatives at each pixel by quantizing the directional derivatives into ternary values to generate Local Ternary Patterns (LTP).
- Orientation selectivity based visual pattern for reduced-reference image quality assessment98
Experimental results on five public databases demonstrate that the proposed RR IQA method has performance consistent with the human perception under a small amount of reference data (only 9 values).
- 96
- 91
- 90
- 87
- Real‐time image smoke detection using staircase searching‐based dual threshold AdaBoost and dynamic analysis87
The authors combine dual threshold AdaBoost with staircase searching technique to propose and implement an image smoke detection method that has a good robustness in terms of early smoke detection and low false alarm rate and can detect smoke from videos with size of 320 × 240 in real time.
- Multi-Scale Frequency Reconstruction for Guided Depth Map Super-Resolution via Deep Residual Network85
- Colour light field image encryption based on DNA sequences and chaotic systems85
This work proposes an encryption scheme for colour LFI based on DNA sequences and chaotic systems and shows that the proposed scheme can be applicable, reliable, and secure enough.
- 84
- 78
- 76
- 76
- Anomaly detection in video sequences: A benchmark and computational model75
A new Large-scale Anomaly Detection (LAD) database is contributed as the benchmark for anomaly detection in video sequences and a multi-task deep neural network is proposed to solve it.
- 71
- 70
- Blind visual quality assessment for image super-resolution by convolutional neural network70
Experimental results show that the proposed metric can obtain better performance than other existing ones in visual quality assessment of image super-resolution, and can be used to estimate the image quality accurately.
- IMFNet: Interpretable Multimodal Fusion for Point Cloud Registration66
A novel attention-fusion module is designed to extract the weighted texture information for descriptor extraction and an interpretable module is proposed to explain the authors' neural network by visually showing the original points contributing to the final descriptor.
- 66
- 65
- 63
- 61
- 59
- 58
- Hyperspectral Image Dataset for Benchmarking on Salient Object Detection57
A hyperspectral salient object detection dataset with a collection of 60 hyperspectrals with their respective ground-truth binary images and representative rendered colour images (sRGB) is released.
- Image salient regions encryption for generating visually meaningful ciphertext image57
This work uses salient regions as important visual information for encryption to obtain ciphertext image for the first time and demonstrates the image salient regions have been largely hidden with the proposed method.
- BSD: Blind image quality assessment based on structural degradation56
This paper proposes a general-purpose blind image quality assessment (BIQA) method by taking account of both the first- and second-order image structures, and extensive experiments manifest the outstanding performance of the proposed method compared to the relevant state-of-the-art BIQA methods.
- 55
- 55
- 54
- 53
- 48
- Distilling Knowledge From Object Classification to Aesthetics Assessment48
This work proposes to distill knowledge on semantic patterns for a vast variety of image contents from multiple pre-trained object classification (POC) models to an IAA model, expecting the combination of multiple POC models can provide sufficient knowledge on various image contents to easier learn to relate various distinct contents to a limited number of aesthetic labels.
- 47
- A Novel Video Salient Object Detection Method via Semisupervised Motion Quality Perception47
A universal learning scheme to obtain a further 3% performance improvement for all state-of-the-art (SOTA) VSOD models and a new concept for mining video frames from the ‘buffered’ testing video stream for constructing a fine-tuning set.
- Joint graph optimization and projection learning for dimensionality reduction47
In order to well exploit and preserve the local structure information of high-dimensional data, a locality constraint is introduced into the proposed JGOPL to discourage a sample from connecting with the distant samples during graph optimization.
- 47
- Stereoscopic image quality assessment by deep convolutional neural network47
Experimental results show that the proposed method can estimate the visual quality of stereoscopic images accurately, which also demonstrate the effectiveness of the proposed two-channel convolutional neural network in simulating the perception mechanism in the HVS.
- 45
- 44
- 43
- 43
- A novel superpixel-based saliency detection model for 360-degree images43
Experimental results on a public eye tracking database of 360-degree images show promising performance of saliency prediction from the proposed superpixel-level saliency detection model by figure-ground law of Gestalt theory.
- 43
- 41
- 40
- A Hybrid Method for Multi-Focus Image Fusion Based on Fast Discrete Curvelet Transform40
A fast discrete Curvelet transform (FDCT)-based technique for multi-focus image fusion to address two problems: texture selection in FDCT domain and block effect in spatial-based fusion.
- 39
- Sub Oriented Histograms of Local Binary Patterns for Smoke Detection and Texture Classification38
This paper proposes sub oriented histograms of LBP for smoke detection and image classification that not only has better performance than existing methods in smoke detection, but also has good performance in texture classification.
- 37
- 37
- 37
- 36
- 36
- 36
- Perceptual Quality Assessment of Omnidirectional Images35
This paper constructs the largest so far subjective omnidirectional image quality database, where several key influential elements are considered, i.e., realistic non-uniform distortion, viewing condition, and viewing behavior, from the user view, and makes the first attempt by using the proposed database to train a convolutional neural network (CNN) for blind omnid directionality image quality assessment.
- Residual dense network for intensity-guided depth map enhancement34
A novel DCNN is proposed to progressively reconstruct the high-resolution depth map guided by the intensity image, where the multi-scale intensity features are extracted to provide guidance for the refinement of depth features as their resolutions are gradually enhanced.
- 33
- 33
- Video saliency detection by gestalt theory32
Experimental results show that a novel adaptive entropy-based uncertainty weighting method designed to fuse spatial and temporal saliency maps to predict the final spatiotemporal saliency map by Gestalt theory can predict visual saliency more accurately than the state-of-art spatiotmporal salency detection algorithms.
- 2AFC Prompting of Large Multimodal Models for Image Quality Assessment31
This work employs the two-alternative forced choice (2AFC) prompting, as 2AFC is widely regarded as the most reliable way of collecting human opinions of visual quality, and introduces three evaluation criteria: consistency, accuracy, and correlation, to provide comprehensive quantifications and deeper insights into the IQA capability of five LMMs.
- 31
- 31
- 31
- 30
- A Novel Long-Term Iterative Mining Scheme for Video Salient Object Detection30
A novel VSOD approach that converts the sequential VSOD, a sequential task, to a data mining problem, i.e., decomposing the input video sequence to object proposals in advance and then mining salient object proposals as much as possible in an easy-to-hard way is proposed.
- 30
- 30
- 30
- 29
- Perceptual Quality Assessment of Omnidirectional Images: A Benchmark and Computational Model29
A large-scale subjective assessment database is built and a novel transformer-based omnidirectional image quality assessment model is proposed to be consistent with the human viewing process, viewing conditions and behaviors are naturally incorporated into the proposed model.
- 27
- 27
- 27
- Saliency-based stereoscopic image retargeting27
Experimental results have shown that both the stereoscopic saliency detection and image retargeting methods can obtain better performance than the existing related methods on the public databases.
- Measuring Perceptual Color Differences of Smartphone Photographs26
This article puts together so far the largest image dataset for perceptual CD assessment, and makes one of the first attempts to construct an end-to-end learnable CD formula based on a lightweight neural network, as a generalization of several previous metrics.
- 25
- Image Quality Assessment in the Modern Age25
This tutorial provides the audience with the basic theories, methodologies, and current progresses of image quality assessment (IQA), and presents in detail the design principles of objective quality assessment models, supplemented by an in-depth analysis of their advantages and disadvantages.
- 25
- 25
- 25
- Saliency-based image retargeting in the compressed domain25
The proposed image retargeting algorithm effectively preserves the objects of attention, efficiently removes the less crucial regions, and therefore significantly outperforms the relevant state-of-the-art algorithms, as demonstrated with the careful analysis and in the extensive experiments.
- 24
- 24
- 24
- 24
- 24
- An efficient and high-quality pansharpening model based on conditional random fields24
The proposed pansharpening model can achieve the best fusion results with high computational efficiency and a total variation regularization is designed to make the gradient of the HRMS image sparse.
- Salient Object Detection on Hyperspectral Images Using Features Learned from Unsupervised Segmentation Task24
Experimental evaluations demonstrated that proposed saliency detection algorithm on hyperspectral images is outperforming state-of-the-arts hyperspectrals saliency models including the original MR based saliency model.
- Blind image quality assessment based on joint log-contrast statistics24
This paper builds a novel NSS model of joint log-contrast distribution to take into account the across space and direction correlations of natural images and provides a new efficient approach to extract quality-aware features as the gradient of log-likelihood on the N SS model, instead of using model parameters directly.
- Performance Evaluation of Visual Tracking Algorithms on Video Sequences With Quality Degradation24
The initial investigation shows that it is still challenging for effective object tracking for existing visual tracking algorithms on video sequences with quality degradation, and proposes a method for robustness measurement of visual tracking (RMVT) algorithms by accuracy rate and performance stability.
- Perceptual quality evaluation for image defocus deblurring24
A quality enhancement module is proposed based on Gray Level Co-occurrence Matrix (GLCM), which is mainly used to measure the loss of texture naturalness caused by deblurring, and demonstrates the effectiveness of the proposed method.
- 23
- 23
- Exposing Semantic Segmentation Failures via Maximum Discrepancy Competition23
This paper exposes failures of existing semantic segmentation methods in the open visual world under the constraint of very limited human labeling effort, and conducts a thorough MAD diagnosis of ten PASCAL VOC semantic segmentsation algorithms.
- Color image quality assessment based on sparse representation and reconstruction residual23
Experiments on public databases demonstrate that the proposed method achieves promising performance in evaluating traditional distortions, and it outperforms the existing metrics when used for quality evaluation of color-distorted images.
- 22
- 22
- 22
- Visual acuity inspired saliency detection by using sparse features22
A new computational model of visual attention based on the relevant characteristics of the Human Visual System and sparse features and the center-bias factor is adopted to enhance the saliency map.
- 21
- 21
- Reduced-reference quality assessment of image super-resolution by energy change and texture variation21
Experimental results demonstrate that the proposed reduced-reference quality assessment metric for image super-resolution (RRIQA-SR) can obtain better performance of quality prediction for SR images than other existing ones, even including some full-reference (FR) metrics.
- Visual structural degradation based reduced-reference image quality assessment20
Experimental results demonstrate that the proposed RR IQA method is highly consistent with the human perception, which uses a small amount of reference data and achieves a promising improvement on quality evaluation accuracy.
- 20
- Max360IQ: Blind omnidirectional image quality assessment with multi-axis attention19
A novel and effective blind omnidirectional image quality assessment (BOIQA) model with multi-axis attention (Max360IQ), which can proficiently measure not only the quality of uniformly distorted omnidirectional images but also the quality of non-uniformly distorted omnidirectional images.
- Omnidirectional Image Quality Captioning: A Large-Scale Database and a New Model19
A novel multitask-derived adaptive feature-tailoring OIQA model named IQCaption360, which is capable of generating a quality caption for an omnidirectional image in a manner of textual template is proposed, which outperforms state-of-the-art methods by a significant margin on the proposed OIQ-10K database.
- Image dehazing based on a transmission fusion strategy by automatic image matting19
A transmission fusion strategy based on automatic image matting for image dehazing that significantly outperforms existing methods and can avoid the difficulty in manually labeling trimaps.
- Perceptual objective quality assessment of stereoscopic stitched images19
A new perceptual objective stereoscopic stitched image quality assessment (S-SIQA) method by considering different distortion types in the existing stitching methods, including color distortion, ghost distortion, structure distortion(shape distortion, information loss), and disparity distortion.
- 19
- Multi-Frame Super-Resolution Reconstruction Based on Gradient Vector Flow Hybrid Field19
A novel multi-frame super-resolution (SR) method developed by considering image enhancement and denoising into the SR processing can effectively suppress both Gaussian and salt-and-pepper noise, meanwhile enhance edges of the reconstructed image.
- 19
- 19
- 19
- Subjective and Objective Quality Assessment of Non-Uniformly Distorted Omnidirectional Images18
A perception-guided OIQA model for non-uniform distortion is proposed by adaptively simulating users' viewing behavior and experimental results demonstrate that the proposed model outperforms state-of-the-art methods.
- Learning a No-Reference Quality Predictor of Stereoscopic Images by Visual Binocular Properties18
A novel no-reference quality assessment metric for stereoscopic images based on monocular and binocular features, motivated by visual perception properties of the human visual system (HVS) named binocular rivalry andbinocular integration is developed.
- 18
- 18
- 17
- 17
- 17
- 17
- Saliency detection by forward and backward cues in deep-CNN17
This paper proposes a top-down saliency model using CNN, a weakly supervised CNN model trained for 1000 object labelling task from RGB images that gives promising results compared to state-of-the-art models in two different datasets.
- 17
- Perceptual Quality Assessment of Virtual Reality Videos in the Wild16
The VR Video Quality in the Wild (VRVQW) database is constructed, containing 502 user-generated videos with diverse content and distortion characteristics, and an objective quality assessment model for VR videos based on pseudocylindrical representation and convolution is developed.
- 16
- A multi-level approach with visual information for encrypted H.265/HEVC videos16
It is found that both encrypting the luma intraprediction model (IPM) and scrambling the syntax element of the DCT coefficient sign can achieve the performance of a distorted video in which there is still residual visual information, while encrypting both of them can implement the intensity of encryption and one cannot gain any visual information.
- GMF: General Multimodal Fusion Framework for Correspondence Outlier Rejection16
Two cross-attention-based fusion layers are proposed to fuse the texture information from paired images and structure information from point correspondences and a convolutional position encoding layer to enhance the difference between $Tokens$ and enable the encoding feature pay attention to neighbor information.
- 16
- 16
- Video saliency detection in the compressed domain16
A video saliency detection model based on feature contrast in the compressed domain is proposed for videos in the uncompressed domain and experimental results show good performance.
- Image Privacy Protection: A Survey15
A systematic review of image privacy protection approaches based on privacy protection goals and defines the attribute known as privacy-sensitive domains and uses it as the primary classification dimension to construct a comprehensive framework for image privacy protection that encompasses various scenarios and privacy objectives.
- 15
- 15
- 15
- 15
- Video Quality Assessment for Online Processing: From Spatial to Temporal Sampling14
An initial attempt is made to design an online VQA model, which is instantiated by as simple as possible a spatial feature extractor, a temporal feature fusion module, and a global quality regression module, and the feasibility of online VQA model is verified by simplifying itself and reducing input.
- 14
- Authenticable medical image-sharing scheme based on embedded small shadow QR code and blockchain framework14
This paper proposes an authenticable medical image-sharing scheme based on an embedded small shadow QR code and a blockchain framework that adopts a hash bit stream for authentication purposes and employs smart contracts to authenticate and restore secret images, thus reducing the local load.
- Revisiting image captioning via maximum discrepancy competition14
A novel method based on maximum discrepancy competition to diagnose existing ICMs and yields several interesting findings, including that using simultaneously low- and high-level object features may be an effective tool to boost the generalization ability for the Transformer based ICMs.
- 14
- 14
- Learning visual saliency from human fixations for stereoscopic images14
A novel stereoscopic saliency detection algorithm by machine learning technique that can predict better saliency results for stereoscopic images than other existing ones is designed.
- 14
- Perceptual Quality Assessment of 360° Images Based on Generative Scanpath Representation13
A unique generative scanpath representation (GSR) is introduced for effective quality inference of 360° images, which aggregates varied perceptual experiences of multi-hypothesis users under a predefined viewing condition and proposes an efficient OIQA computational framework by learning the quality maps of GSR.
- 13
- Meta-Point Learning and Refining for Category-Agnostic Pose Estimation13
This work maintains learnable embeddings to capture inherent information of various keypoints, which interact with image feature maps to produce meta-points without any support, which could serve as meaningful potential keypoints for CAPE.
- Multitask Auxiliary Network for Perceptual Quality Assessment of Non-Uniformly Distorted Omnidirectional Images13
A multitask auxiliary network for non-uniformly distorted omnidirectional images, where the parameters are optimized by jointly training the main task and other auxiliary tasks, and these auxiliary sub-networks contribute to improve the performance of the proposed model.
- Light field image encryption based on spatial-angular characteristic13
This paper proposes a novel scheme for LF image encryption based on the internal spatial-angular characteristic of LF image, and uses the discrete fractional Fourier transform to process the spatial domain of the LF image.
- Analysis of the Impact of Medical Features and Risk Prediction of Acute Kidney Injury for Critical Patients Using Temporal Electronic Health Record Data With Attention-Based Neural Network13
An attention-based temporal neural network approach was employed in this study for AKI prediction and for the analysis of the impact of medical features from temporal electronic health record (EHR) data of patients before AKI diagnosis.
- 13
- Ordinal preserving matrix factorization for unsupervised feature selection13
A novel algorithm called Ordinal Preserving Matrix Factorization, which incorporates matrix factorization, ordinal locality structure preserving and inner-product regularization into a unified framework, is proposed for feature selection and can obtain competitive performance compared to the existing state-of-the-art unsupervised feature selection approaches.
- 12
- 12
- Blind quality assessment for tone-mapped images based on local and global features12
Experimental results indicate that the proposed metric can get better performance in predicting the visual quality of tone-mapped images than the state-of-the-art methods.
- 12
- 12
- 12
- Viewport-Unaware Blind Omnidirectional Image Quality Assessment: A Flexible and Effective Paradigm11
This article presents a flexible and effective paradigm, which is viewport-unaware and can be easily adapted to 2D plane image quality assessment (2D-IQA), which achieves competitive performance with low complexity against other state-of-the-art models and its adaptive capacity to 2D-IQA is verified.
- 11
- 11
- CD-iNet: Deep Invertible Network for Perceptual Image Color Difference Measurement11
The resulting metric, referred to as CD-iNet, is end-to-end optimized on color patch datasets and image datasets simultaneously and can produce competitive local CD maps without requiring dense supervision and be robust against geometric distortions.
- 11
- DevsNet: Deep Video Saliency Network using Short-term and Long-term Cues11
A novel deep learning framework of saliency detection for video sequences, namely Deep Video Saliency Network (DevsNet), which mainly consists of two components: 3D Convolutional Network (3D-ConvNet) and Bidirectional convolutional Long-Short Term Memory Network (B-ConVLSTM).
- 11
- Subjective quality evaluation of compressed digital compound images11
Experimental results indicate that the existing IQA methods are limited in evaluating visual quality of DCIs.
- Multi-level perturbations in image and feature spaces for semi-supervised medical image segmentation10
- Visual Security Index Combining CNN and Filter for Perceptually Encrypted Light Field Images10
A new database of perceptually encrypted LF images is constructed, called the PE-SLF, which is 2.6 times as big as the existing largest perceptual encrypted image database, and a novel visual security index (VSI) model is proposed by taking into full consideration the intrinsic spatial-angular characteristics of the LF images and the outstanding capabilities of CNN in feature extraction.
- Unsupervised stereoscopic image retargeting via view synthesis and stereo cycle consistency losses10
An unsupervised stereoscopic image retargeting network (USIR-Net), which consists of a content consistency term and a disparity consistency term, is proposed to preserve the structure information and prevent binocular disparity inconsistency and achieves superior performance compared with state-of-the-art methods.
- Video Saliency Detection by 3D Convolutional Neural Networks10
A novel and effective approach for salient object detection for video sequences based on 3D convolutional neural networks and a 3D deconvolutional network to combine the spatiotemporal features to predict the final saliency map for video sequence.
- 10
- 9
- Spatial Quality Oriented Rate Control for Volumetric Video Streaming via Deep Reinforcement Learning9
- Lesion‐aware network for diabetic retinopathy diagnosis9
A CNN‐based DR diagnosis network with attention mechanism involved, termed lesion‐aware network, to better capture lesion information from imbalanced data, which outperforms the mainstream methods with an area under curve of 0.967 in DR screening and increases the overall average precision.
- Dynamic proposal sampling for weakly supervised object detection9
The primary contribution of this work is to improve weakly supervised detection (WSD) with a dynamic proposal sampling (DPS) strategy that collects purified positive training samples by progressively removing confident background clutters, and selects discriminative negative samples by mining class-specific hard proposals.
- 9
- 9
- Diffusion-Based Facial Aesthetics Enhancement With 3D Structure Guidance8
The Nearest Neighbor Structure Guidance based on Diffusion (NNSG-Diffusion), a diffusion-based FAE method that beautifies a 2D facial image with 3D structure guidance, is proposed, which is superior to previous relevant methods in enhancing facial aesthetics while preserving facial identity.
- 8
- Improving Image Aesthetic Assessment via Multiple Image Joint Learning8
This work presents a novel Multiple Image joint Learning Network (MILNet) to mimic this natural process that people consider similar experiences to improve the consistence of the voting result of the IAA methods.
- 8
- 8
- 8
- Exploiting entropy masking in perceptual graphic rendering8
This paper detects unstructured regions by extracting local patches from each pixel and its neighboring pixels, and comparing the similarity between the local patches of the center pixel and the neighboring pixel, and exploits entropy masking in perceptual rendering.
- 8
- 7
- 7
- Robustness Analysis of Pedestrian Detectors for Surveillance7
This paper designs a method to evaluate detection stability and a robustness measure called robustness quadrangle, which can be adopted to the visualize detection accuracy of pedestrian detection algorithms on high-quality video sequences and stability with video quality degradation.
- 7
- 7
- 7
- A Novel Spatial Pooling Strategy for Image Quality Assessment7
This work proposes a novel spatial pooling strategy for image quality assessment through statistical analysis of the quality map and indicates that the overall image quality is sensitive to the quality distribution.
- 7
- 6
- 6
- CMF-IoU: Multi-Stage Cross-Modal Fusion 3D Object Detection With IoU Joint Prediction6
An iterative voxel-point aware fine grained pooling module is introduced, which captures the spatial information from LiDAR points and textural information from pseudo points in the proposal refinement stage to effectively address the challenge of aligning 3D spatial and 2D semantic information.
- 6
- Weak-shot Keypoint Estimation via Keyness and Correspondence Transfer6
This work proposes to transfer keyness and correspondence, which essentially belong to comparing entities and thus are class-agnostic and class-wise transferable, which can activate the keypoints of novel classes by reinforcing the consistency between two paired images.
- 6
- Comprehensive Visual Grounding for Video Description6
A comprehensive visual grounding network to improve video captioning, by explicitly linking the entities and actions to the visual clues across the video frames, which encourages the attention mechanism to focus on informative spatial areas across video frames.
- 6
- Integrating Large Pre-trained Models into Multimodal Named Entity Recognition with Evidential Fusion6
This work models the distribution of each modality as a Normal-inverse Gamma distribution, and fuses them into a unified distribution with an evidential fusion mechanism, enabling hierarchical characterization of uncertainties and promotion of prediction accuracy and trustworthiness in MNER.
- 6
- 6
- 6
- Objective quality assessment of synthesized images by local variation measurement6
Experimental results show that the performance of the proposed algorithm is competitive with state-of-the-art FR and no-reference image quality assessment metrics.
- 6
- 6
- Saliency detection in computer rendered images based on object-level contrast6
This work proposes a novel graphic saliency detection method to detect visually salient objects in images rendered from 3D geometry models and can obtain much better performance in salient object detection.
- Bottom-Up Saliency Detection Model Based on Amplitude Spectrum6
The novel saliency measure for image patches by using amplitude spectrum of QFT proves promising, as the experiment results show that this saliency detection model performs better than the relevant existing models.
- From Global to Granular: Revealing IQA Model Performance via Correlation Surface5
Granularity-Modulated Correlation (GMC) is proposed, which provides a structured, fine-grained analysis of IQA performance and maps correlation values as a joint function of MOS and <inline-formula><tex-math notation="LaTeX">$|\Delta$</tex-math><alternatives><mml:mrow><mml:mo>|</mml:mo>
- A Perceptually Optimized and Self-Calibrated Tone Mapping Operator5
A two-stage neural network-based TMO that is self-calibrated and perceptually optimized that produces images with consistently better visual quality while ranking among the fastest local TMOs.
- Recurrent Feature Mining and Keypoint Mixup Padding for Category-Agnostic Pose Estimation5
A FGSA mining module based on deformable attention mechanism, which recurrently mines FGSA features from both support and query images, and proposes to use mixup keypoints to pad various classes to a unified keypoint number, which could provide richer supervision than the zero padding used in existing works.
- 5
- 5
- Computational Analysis of Degradation Modeling in Blind Panoramic Image Quality Assessment5
This article presents a sufficient computational analysis of degradation modeling in BPIQA to thoroughly explore the easy-database issue, and finds that easy databases narrow the gap between the performance of BPIQA and BIQA models, which is unconducive to the development of BPIQA.
- 5
- 5
- 5
- 5
- 5
- 5
- Objective Quality Assessment of Screen Content Images by Structure Information5
Experimental results show that the proposed method can obtain better performance of visual quality prediction of SCIs than other existing ones.
- No-reference Image Quality Assessment Based on Structural and Luminance Information5
A novel no-reference IQA metric, based on the analysis of structural degradation and luminance changes, is introduced, which encodes the image structural information as the local binary pattern LBP distribution.
- 5
- 5
- A Novel Objective Quality Assessment for Super-Resolution Images5
Experimental results show that the proposed objective quality assessment method can be used in the real applications with the original high-resolution images unavailable and an overall evaluation of visual quality of the super-resolution image can be computed.
- Detecting Malicious Concepts Without Image Generation in AI-Generated Content (AIGC)4
The proposed Concept QuickLook is the first systematic work to incorporate malicious concept detection into research, which performs detection based solely on concept files without generating any images, and defines malicious concepts and design two operational modes for detection: concept matching and fuzzy detection.
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- Abnormal Event Detection Based on Saliency Information4
Experimental results show that the proposed abnormal event detection method can obtain much better performance than the existing ones over the public video database.
- Viewport-Unaware Blind Omnidirectional Image Quality Assessment: A Unified and Generalized Approach3
This paper experimentally finds that BOIQA can be formulated as a blind (2D planar) image quality assessment (BIQA) problem, and presents a new BOIQA approach, which has three merits: viewport-unaware, unified, and generalized, which shows better generalizability against other competitors.
- 3
- 3
- 3
- 3
- Inter-modality feature prediction through multimodal fusion for 3D shape defect detection3
A lightweight inter-modality feature prediction framework that effectively utilizes multimodal fused features from the inputs of RGB, depth and point clouds for efficient 3D shape defect detection and eliminates the need for large memory banks or pixel-level reconstructions is proposed.
- 3
- 3
- 3
- 3
- 3
- Asymmetrically distorted 3D video quality assessment: From the motion variation to perceived quality3
- 3
- 3
- A general effective rate control system based on matching measurement and inter-quantizer3
A general effective rate control system based on matching measurement and inter-� quantizer (RCMMAIL) for actual video, which includes a novel Frame-Level Rate Control algorithm based on Matching Measurement (FRCMM) and a novel inter- � quantizer that can solve the problem of the uncertain parameters in the current � model, improving the accuracy and reducing computational complexity simultaneously.
- 3
- 3
- 3
- Blind Omnidirectional Image Quality Assessment: Embracing the Magic Power of Multimodal Large Language Models2
A family of MultiModal BOIQA models based on the embeddings in the built vision-language space are designed, comprehensively investigating the effectiveness of text features, visual features, and their interaction in capturing quality degradation of OIs.
- 2
- 2
- 2
- 2
- 2
- 2
- Enhancing Incomplete Multi-modal Brain Tumor Segmentation with Intra-modal Asymmetry and Inter-modal Dependency2
A novel approach that enhances the BTS model from two perspectives is proposed, introducing a pre-training stage that generates a diverse pre-training dataset covering a wide range of different combinations of tumor shapes and brain anatomy and a post-training stage that enables the model to reconstruct missing modalities in the prediction results when only partial modalities are available.
- Real-time Wing Deformation Simulations for Flying Insects2
This paper presents an efficient skeleton-driven model specifically designed to real-time simulate realistic wing deformations across a wide range of flying insects, and demonstrates the effectiveness, robustness, and adaptability of this model.
- 2
- Dynamic Attention Vision-Language Transformer Network for Person Re-identification2
A novel image-text dynamic attention module (ITDA) is designed to promote unbiased feature integration by dynamically assigning the importance of image and text representations and an adapter module is adopted to address the domain gap between pretraining datasets and ReID datasets.
- 2
- 2
- A Practical Method for Butterfly Motion Capture2
This paper proposes a practical yet effective, optical marker-based approach to capture and process the detailed motion of a flying butterfly, and created and will release a first-of-its-kind butterfly motion capture dataset to research community.
- 2
- 2
- Improved Salient Object Detection Based on Background Priors2
A novel saliency detection model is proposed by computing the contrast between superpixels with background priors and introducing a refinement method to address the problem in existing studies.
- Application-specific visual quality assessment2
Different types of V QA metrics for specific applications are introduced and the future trends of VQA for specific applied applications are described.
- 2
- Infrared-Privileged UAV Detection via Cross-Modal Vector-Quantization1
This work explores a new learning paradigm using the IR modality as privileged information, and proposes to quantize the IR features and hallucinate the codebook-indices based on RGB features, which is more effective and robust than directly hallucinating features.
- 1
- SinMDGan: A Hybrid Deep Learning Framework for Single Motion Synthesis Using Diffusion‐GAN Models1
SinMDGan is introduced, a hybrid deep learning framework for single‐motion synthesis that leverages a Diffusion‐GAN model, and employs a single generator‐discriminator pair, utilizing different diffusion time steps to synthesize novel and diverse motions from a single short sequence.
- A Dual-Protection Method for 3D Object Security and Copyright: Watermark Embedding During Decryption1
- MERD-360VR: A Multimodal Emotional Response Dataset from 360° VR Videos Across Different Age Groups1
This publicly available multimodal dataset provides a valuable resource to facilitate numerous future efforts on VR-based affective computing research and on tailoring VR content to diverse audiences based on emotional and demographic profiles.
- Foreground-aware transformer network for person re-identification1
The proposed foreground-aware transformer network isolate the foreground by minimizing the impact of background interference and introduce a foreground-aware loss function, which directs the attention of networks toward the primary foreground information in the image, optimizing its ability to identify pedestrians.
- Make Identity Unextractable yet Perceptible: Synthesis-Based Privacy Protection for Subject Faces in Photos1
This paper proposes the first synthesis-based method dedicated to subject faces, i.e., PerceptFace, which can make identity unextractable yet perceptible, and provides a public API of PerceptFace, which has great potential to become a practical anti-FR tool.
- 1
- Texture‐Aware Network for Enhancing Inner Smoke Representation in Visual Smoke Density Estimation1
It is argued that pixels within smoke exhibit a probabilistic relationship with both smoke and background, necessitating density estimation to enhance the representation of internal structures within the smoke, and proposes enhancements across the entire network.
- 1
- 1
- 1
- 1
- 1
- 1
- Saliency-based artificial object detection for satellite images1
The top-down saliency is integrated with bottom-up saliency to obtain the saliency map in satellite images and the method is compared to several state-of-the-art saliency detection models and demonstrates the superior performance in artificial object detection for satellite images.
- Moving Object Tracking with Structure Complexity Coefficients1
An Observation Dependent Hidden Markov Model (OD-HMM) framework is designed where the observation dependency between neighboring frames is considered comparing with the standard HMM based tracking framework and a novel outlier removing method in appearance model updating to avoid error accumulation is presented.
- 1
- 1
- 1
- –
- FreDeGS: frequency-guided 3D gaussian splatting for scene deblurring–
FreDeGS is proposed, a frequency-guided 3DGS deblurring framework that adopts a physical convolution synthesis baseline to decouple static scene geometry from complex blur and introduces the Statistics-Wavelet Gated Multi-scale Adapter (SWGMA), leveraging channel statistics and spatial wavelet energy to enhance high-frequency perception beyond purely spatial constraints.
- –
- Evidential Robust Feature Learning for Generalized Few-Shot Segmentation–
This work introduces an evidential learning approach that simultaneously improves feature representation robustness and reduces model uncertainty and leverages both general-purpose features from segmentation-centric large pre-trained foundation models and low-level structural features with a physics-informed constraint to improve representations, especially for novel classes.
- –
- –
- –
- –
- –
- –
- –
- Toward Realistic Low-Light Image Enhancement via ISP-Driven Data Modeling–
This paper proposes an image signal processing (ISP)-driven data synthesis pipeline that addresses challenges by generating scalable paired training data and applies it to existing LLIE methods, demonstrating its impact on improving their practicality and generalizability in real-world applications.
- –
- –
- –
- TaylorMoDe-GS: Taylor-Driven Gaussian Splatting Motion Model for Multi-View Dynamic Scene Deblurring–
TaylorMoDe-GS, the first 3DGS framework tailored for multi-view dynamic object deblurring, shifts the modeling paradigm from displacement fitting to velocity driven modeling, and introduces a neural Peano remainder network to compensate for high frequency non-linear dynamics.
- –
- –
- Event-Guided Super-Resolving Blurry Image via Asymmetric Integral Driven Consistency–
The proposed SE-SRB introduces a self-supervised learning paradigm based on asymmetric integral driven consistency, which enforces temporal coherence between predictions derived from RGB and asynchronous event streams at different time points.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- DR-SISM: A dual reversible secret image sharing mechanism–
This paper proposes a dual reversible secret image sharing mechanism, called DR-SISM, which includes reversible hiding stage and reversible sharing stage that can simultaneously achieve the reversible recovery of secret image and additional data embedded.
- –
- –
- A Perceptually Optimized and Self-Calibrated Tone Mapping Operator–
This paper develops a two-stage neural network-based HDR image TMO that is biologically-inspired, computationally efficient, and perceptually optimized, and fully automatic to tone map uncalibrated HDR images.
- –
- –
- –
- Omnidirectional Images as Moving Camera Videos.–
This paper first transforms an omnidirectional image to several video representations using viewing behavior of different users, and then leverage the recent advances in full-reference 2D image/video quality assessment to compute the perceived quality of the panorama.
- –
- Blind Quality Assessment for DIBR-Synthesized Images Based on Chromatic and Disoccluded Information–
A novel method for no reference image quality assessment by considering chromatic and disoccluded information, inspired by the characteristic of the human visual system (HVS) and the synthesized image-specific distortion type, which shows improved performance on IRCCyN/IVC DIBR and MCL-3D databases.
- The Application of Reasonableness Evaluation Model Based on Coefficient of Variation and Normal Distribution in Advanced Mathematics–
A comprehensive evaluation model is built, including the relevance, importance, difficulty, extensibility and enlightenment of knowledge points in the database, to formulate a reasonable Question program for advanced mathematics BI.
- –
- –
- –
- –
- –
- –
- Image Interpolation Based on Saliency Information–
Experimental results show that the proposed method improves the perceptual quality of the salient regions during image interpolation and can obtain better performance than existing related image interpolations methods.
- An Effective Video Saliency Detection Model Based on Human Visual Acuity and Spatiotemporal Cues in Cloud Systems–
Experimental results demonstrate that the performance of the proposed video saliency detection model in cloud systems is better than other existing ones.
- –
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 2026-10-12. 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.