Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks115 from public data
- 541
- 168
- 154
- A high capacity lossless data hiding scheme for JPEG images150
Experimental results show that the proposed high-capacity reversible data hiding method achieves both high capacity and high image quality.
- 140
- Beyond the Prior Forgery Knowledge: Mining Critical Clues for General Face Forgery Detection126
This work proposes a novel Critical Forgery Mining framework, which can be flexibly assembled with various backbones to boost their generalization and robustness performance, and designs a novel progressive learning controller to guide the model to focus on principal feature components, enabling it to learn critical forgery features in a coarse-to-fine manner.
- Learning Meta Pattern for Face Anti-Spoofing92
A learnable network to extract Meta Pattern (MP) and a two-stream network to hierarchically fuse the input RGB image and the extracted MP by using the proposed Hierarchical Fusion Module (HFM) are proposed.
- Wavelet domain adaptive visible watermarking91
A wavelet domain visible watermarking is proposed, adaptively determined by the effect of luminance and local spatial characteristics to make the degree of marking applied variable according to the features of host images.
- Rethinking Vision Transformer and Masked Autoencoder in Multimodal Face Anti-Spoofing74
The modality-asymmetric masked autoencoder is proposed for multimodal FAS self-supervised pre-training without costly annotated labels and is able to learn more intrinsic task-aware representation and compatible with modality-agnostic downstream settings.
- An improved VLC-based lossless data hiding scheme for JPEG images66
A lossless data hiding scheme which directly embeds data into the bitstream of JPEG images is presented and can achieve a significant improvement of embedding capacity.
- 64
- Robust Clothing-Invariant Gait Recognition50
A random subspace method (RSM) framework for clothing-invariant gait recognition by combining multiple inductive biases for classification is proposed and can achieve promising performance even for small size training set.
- 47
- S-Adapter: Generalizing Vision Transformer for Face Anti-Spoofing With Statistical Tokens46
A generalized FAS method under the Efficient Parameter Transfer Learning (EPTL) paradigm, where the pre-trained Vision Transformer models are adapted for the FAS task, and a novel Statistical Adapter (S-Adapter) that gathers local discriminative and statistical information from localized token histograms is proposed.
- Self-Supervised 3D Action Representation Learning With Skeleton Cloud Colorization44
This paper addresses self-supervised 3D action representation learning for skeleton-based action recognition and designs a novel skeleton cloud colorization technique that is capable of learning spatial and temporal skeleton representations from unlabeled skeleton sequence data.
- 44
- 41
- 40
- 37
- Rehearsal-Free Domain Continual Face Anti-Spoofing: Generalize More and Forget Less36
This paper proposes the first rehearsal-free method for Domain Continual Learning of FAS, which deals with catastrophic forgetting and unseen domain generalization problems simultaneously and proposes the Proxy Prototype Contrastive Regularization (PPCR) to constrain the continual learning with previous domain knowledge from the proxy prototypes.
- Using Graph Neural Networks to Improve Generalization Capability of the Models for Deepfake Detection35
This paper proposes a graph neural network-based paradigm that casts deepfake detection as a graph binary classification problem and proposes a dual-branch network to extract node features from both RGB images and their color difference images via the Transformer-based trainable node encoder module (TNEM).
- Exposing image forgery through the detection of contrast enhancement35
The inter-channel correlation introduced by color image interpolation is revealed, and a metric is constructed to measure these correlations, which are useful in distinguishing the original and contrast enhanced images.
- 33
- Gait recognition based on the golden ratio31
The golden ratio-based segmentation method is proposed to reduce the influence of clothing and outperforms other approaches, achieving a 94.76 % recognition rate in various clothing conditions and a rate of 91.53 % when bags are being carried.
- 30
- 30
- 29
- Multi-Label-Based Similarity Learning for Vehicle Re-Identification29
This paper proposes multi-label-based similarity learning (MLSL) for vehicle re-identification obtaining an efficient deep-learning-based model that derives robust vehicle representations and proves the superiority of the model over multiple state-of-the-art methods on the three mentioned datasets.
- 27
- 24
- 24
- Auto-Zooming CNN-Based Framework for Real-Time Pedestrian Detection in Outdoor Surveillance Videos23
A fast, lightweight, and auto-zooming-based framework for small pedestrian detection which can be run on low computational resources is proposed and compared with different single-step customized pedestrian detectors as well as the two-step detector faster R-CNN.
- Asymmetric Modality Translation for Face Presentation Attack Detection21
A novel framework based on asymmetric modality translation for face presentation attack detection in bi-modality scenarios is proposed, which is effective in detecting various types of attacks and achieves state-of-the-art performance under different evaluation protocols.
- An efficient reversible data hiding scheme using prediction and optimal side information selection20
Experimental results demonstrate the superior performance of the proposed HS based reversible data hiding scheme in the terms of embedding capacity and stego-image quality.
- 20
- Deepfake detection based on cross-domain local characteristic analysis with multi-domain transformer19
- 19
- Reversible data hiding algorithm for vector digital maps19
A difference histogram-based reversible data hiding algorithm was proposed for authentication and covert communications of vector maps that has better embedding rate, better quality of the output image, and more accurate capacity control capability.
- Image Fusion Based Visible Watermarking Using Dual-Tree Complex Wavelet Transform19
This work proposes a new framework of visible watermarking based on image fusion, a common technique used in combining images acquired from different modalities, and addresses the problem of low-pass subband watermark embedding, which is a special issue of visibleWatermarking.
- An Improved Ship Classification Method Based on YOLOv7 Model with Attention Mechanism17
A YOLOv7-residual convolutional block attention module (YOLO v7-RCBAM) is proposed by combining the Convolutional attention mechanism and residual connections to the YOLov7 to accelerate the training speed and enhance the information relevance of channel dimensional features.
- ENF Based Video Forgery Detection Algorithm17
A new ENF-based time domain video forgery detection algorithm is proposed to solve the problem of data lack and has good performance in detecting forgery videos with varying degrees of deletion, duplication and insertion of frames.
- Audio forensic authentication based on MOCC between ENF and reference signals17
This paper proposes a new audio authenticity detection algorithm based on the max offset for cross correlation (MOCC) between the extracted ENF (Electric Network Frequency) signal and the reference signal that can detect not only audio forgery but also the edited region andaudio forgery type.
- Analyzing temporal coherence for deepfake video detection16
Thorough experiments on large-scale datasets reveal that the proposed paradigm surpasses current approaches and stays effective when detecting unseen sorts of deepfake videos.
- An HEVC steganalytic approach against motion vector modification using local optimality in candidate list16
This is the first work of video steganalytic method exploiting the new features of HEVC format, and significantly boosts the detection accuracy of the index modification method that is exclusive to HEVC videos.
- Fast source camera identification using matching signs between query and reference fingerprints16
This paper uses global information derived from the relationship between the query fingerprint/digest and the reference fingerprints/digests in the database to guide fast search, which can provide more accurate and robust clues for the selection of candidate matching database fingerprints.
- 16
- 16
- 15
- 13
- Revisiting Label Smoothing Regularization with Knowledge Distillation13
A new teacher-free framework LSR-OS-TC is developed, which decomposes the Tf-KD method into two components: model Output Smoothing (OS) and Teacher Correction (TC), and shows that L SR-OS can improve LSR performance independently with no extra computational cost, especially on several deep neural networks where LSR is ineffective.
- 13
- 12
- 12
- Audio Forgery Detection Based on Max Offsets for Cross Correlation between ENF and Reference Signal11
A new method based on the max offset for cross correlation between the extracted ENF and the reference signal can determine whether the audio signal in question was digitally edited as well as the location at which the editing manipulation occurs.
- An Improved Fingerprinting Algorithm for Detection of Video Frame Duplication Forgery11
Experimental results show that the proposed algorithm overall outperforms three related duplication forgery detection algorithms in terms of computational efficiency, detection accuracy and robustness against common video operations like compression and brightness change.
- Source Camera Identification Issues11
This work first builds a sample camera classifier using a combination of popular image features, and then reveals its deficiency, and suggests for the design of robust camera classifiers are given.
- 10
- Analysis and Comparison of Typical Reversible Watermarking Methods10
This paper focuses on analyzing and comparing these three reversible watermarking methods and covers several aspects including data hiding capacity, image quality, capacity resilience and control, computational complexity, security, and blind data extraction.
- 10
- 9
- Exploring varying color spaces through representative forgery learning to improve deepfake detection9
- LIS-Net: An end-to-end light interior search network for speech command recognition9
A novel neural network, called Light Interior Search Network (LIS-Net), is presented that works with raw speech signal and is validated on Google Speech Commands benchmark speech datasets, demonstrating a significant improvement of accuracy and processing time in comparison with other state-of-the-art techniques.
- Channel Response Based Multi-Feature Audio Splicing Forgery Detection and Localization9
An audio splicing detection and localization method based on channel response multi-feature based on Gaussian Mixture Model on RASTA-MFCC coefficients to blind channel estimation improves the detection rate of forgery and its localization.
- Hybrid Domain Meta-Learning Network for Face Forgery Detection and Localization in Deepfakes8
A hybrid domain meta-learning network (HDMNet) for face forgery detection and manipulation localization and the effectiveness and good generalization ability of this method is validated compared with several state-of-the-art methods.
- Automatic de-noising and recognition algorithm for drilling fluid pulse signal8
Wavelet forced de-noising algorithm is suitable for denoising of unsteady drilling fluid pulse signal, including baseline drift rectification and two-stage de- noising processing of frame synchronization signal and instruction signal.
- A Novel Online Bayes Classifier8
VIGO algorithm is presented, a novel online Bayesian classifier for both binary or multiclass problems that can be updated in a minibatch of an arbitrary size which makes it robust with noise data.
- An adaptive system for gait recognition in multi-view environments8
The avoidance of searching whole multi-view database as well as the competitive performance indicate that the proposed method is practical for gait recognition in real world surveillance scenarios.
- Building fingerprints with information from three color bands for source camera identification8
This work explores the way to build fingerprints based on three color bands of an image to better reflect the intrinsic characteristics of the camera.
- 8
- Hyperbolic Face Anti-Spoofing6
This work proposes to learn richer hierarchical and discriminative spoofing cues in hyperbolic space, and designs a multimodal FAS framework with Euclidean multi-modal feature decomposition andhyperbolic multimodals feature fusion&classification.
- 6
- An Improved Algorithm for Camera Model Identification Using Inter-channel Demosaicking Traces6
An improved algorithm to use inter-channel demosaicking/color interpolation traces for identifying the source camera model of a test image by extracting two variance maps and taking the shape and texture features of these two maps for camera model identification.
- 5
- Learnable Information-Preserving Image Resizer for Face Forgery Detection5
Experimental results show that the proposed simple yet effective learnable information-preserving resizer outperforms other resizers and consistently enhances the detection performance of tested detectors, confirming the effectiveness of the proposed resizer.
- 5
- 5
- 5
- CBC-Based Synthetic Speech Detection5
By combining CQT, block transform (BT) and discrete cosine transform (DCT), a new feature is obtained, namely, constant-Q block coefficients (CBC), which is superior to other front-ends features that have been benchmarked on ASVspoof 2015 evaluation set in terms of equal error rate (EER).
- WolfPath: Accelerating Iterative Traversing-Based Graph Processing Algorithms on GPU5
A graph preprocessing algorithm is developed in this work to convert any graph into the format of the Layered Edge list, which achieves significant speedup over the state-of-art GPU-based in-memory and out- of-memory graph processing frameworks.
- Study of Hydropower Units Fault Diagnosis based on Bayesian Network Noisy Or Model5
The network is constructed by applying experts' knowledge in determining the probability of each node and taking the characteristic nodes as binary nodes and the influence of random combinations of multiple nodes on the outcome is calculated to determine the possibility of occurrence of a certain fault.
- Deepfake Videos Detection Based on Image Segmentation with Deep Neural Networks4
This article attempts to solve the problem of fake faces detection with the way of image splicing detection by using a neural network borrowed from image segmentation for predicting the tampered face area from which a tampering mask is obtained through denoising and thresholding the probability map.
- Modelling and developing conflict-aware scheduling on large-scale data centres4
This work model distributed scheduling and resource conflict using the game theory and conduct the quantitative analysis about scheduling cost and job performance, and develops the conflict-aware scheduling strategies to reduce the schedulingcost and improve job performance.
- Further studies on forensic features for source camera identification4
A published camera classifier is implemented and the performance of the classifier on images under shearing, histogram equalization, and contrast-stretching operations is investigated and the impact of camera databases of different sizes is probed.
- Evaluating and Enhancing Face Anti-Spoofing Algorithms for Light Makeup: A General Detection Approach3
A general face anti-spoofing algorithm specifically designed for light makeup faces is introduced, which includes a makeup augmentation module, a batch channel normalization module, a backbone network updated via the Exponential Moving Average method, an asymmetric virtual triplet loss module, and a nearest neighbor supervised contrastive module.
- 3
- 3
- Image Inpainting Technology Based on Image Decomposition3
Improved TV model is applied to structure part of image to effectively avoid step effect in TV model on smooth region and improved exemplar-based inpainting technology extracts structure information and texture information from the image.
- Reversible Data Hiding Using Prediction Error Values Embedding3
The experimental results show that the proposed even- and odd-number-based embedding method can achieve satisfactory embedding capacity and image quality.
- The Design and Application of DWT-Domain Optimum Decoders3
An iterative, Monte-Carlo algorithm is presented that tries to ensure watermarks are embedded with a specified value of this robustness estimate, and is found to successfully embed robust, 129-bit watermarks in 368 × 240 images.
- Learning deep forest for face anti-spoofing: An alternative to the neural network against adversarial attacks2
A novel solution for FAS against adversarial attacks, leveraging a deep forest model based on local binary patterns (LBP) as the model input, replacing the grained-scanning mechanism used in the traditional deep forest model.
- 2
- 2
- Three-dimensional Region Forgery Detection and Localization in Videos2
This research proposed a new method, which can be used for detecting and locating 3-D duplicated regions in videos based on the phase-correlation of3-D regions residual more efficiently, and proved that the proposed method is efficient and robust for detecting small 3- D regions duplication and frame sequences duplication.
- Gait recognition under carrying condition: a static dynamic fusion method2
This paper proposes a novel dynamic feature template for classification on the large USF (University of South Florida) database by adopting a static/dynamic fusion strategy and achieves significant improvements when compared with other classic algorithms.
- Exploring Static–Dynamic ID Matching and Temporal Static ID Inconsistency for Generalizable Deepfake Detection1
Results show that the proposed novel Deepfake video detection approach based on identity anomalies can detect not only face replacement but also face reenactment, and also has better detection performance over the state‐of‐the‐art methods on unknown datasets and has superior robustness against compression.
- 1
- Study on performance of typical source camera classification algorithms1
Using Support Vector Machines (SVM), this paper analyzed the performance and robustness of source camera classification algorithms, and compared the detection accuracy for unprocessed images with that for processed images, and investigated the robustity of different types of image features.
- Issues on Selecting Image Features for Robust Source Camera Identification1
This work demonstrates that robustness should also be considered for classifiers which aim at real-world tasks, and reveals what impact the reduced feature subset will have on the robustness of camera classifiers.
- Survey on techniques of digital multimedia forensics1
The authors focused on introducing typical algorithms and pointing out the main problems in the current research and suggested the urgent topics for the future research on digital multimedia forensics.
- Improved algorithm based on TV model for image inpainting1
Improved image inpainting algorithm combines isotropic with anisotropic diffusion and uses the regional frequency difference to realize different iterative equation on the different regional.
- Reversible data hiding giving priority to the use of edges and textures1
The experiments prove that the proposed new difference expansion (DE) embedding based reversible data hiding algorithm can produce visual image quality more suitable for human visual characteristics.
- Jointly learning and training: using style diversification to improve domain generalization for deepfake detection–
A generalizable deepfake detection system to implement style diversification by alternately learning the domain generalization (DG)-based detector and the stylized fake face synthesizer and demonstrates that this method outperforms the state-of-the-art competitors not only in intra-domain tests but particularly in cross-domain tests.
- –
- –
- Well Depth Monitor Software Design Based On Laser Distance Measuring Device–
By the well depth measurement system software based on laser distance measuring device, the application of this system contributes to the drilling operation and shortens the drilling period, and reduces the system error and improved the accuracy of measurement.
- –
- –
- –
- –
- A contourlet transform-based steganalysis algorithm–
A universal steganalysis method was proposed by using the superior property of the contourlet transform with representation of an image to have the superion discriminative performance for most of steganography methods.
- Source camera identification technique using large components of imaging sensor pattern noise–
A new method for source camera identification was proposed based on using large components of imaging sensor pattern noise to improve the performance of the correlation-based detector and greatly decrease the computational load for detection.
- Source camera identification schemes with color image information–
Through analyzing the characteristics of interpolation algorithms in the mosaic image generation process, three color image-based source camera identification schemes were proposed and experimental results show that the new method has better identification capability and needs fewer image pixels.
- Curve watermarking technique for protecting copyright of digital maps–
This watermarked curve suffers smaller embedding distortions and is robust against common geometrical distortions, collusions, and printing-scaling attacks.
- –
- New reversible data hiding algorithm using difference expansion embedding–
Based on integer Haar wavelet transform, a new reversible data hiding algorithm using the horizontal and vertical difference images for difference expansion yields better image quality under the same embedding rate.
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.