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
- Honghai LiuSuggested from co-authorship
- David J BrownSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works228 from public data
- Getting to know low-light images with the Exclusively Dark dataset901
It is found that the effects of low-light reaches far deeper into the features than can be solved by simple "illumination invariance'", and it is hoped that this analysis and the Exclusively Dark dataset can encourage the growth in low-lights domain researches on different fields.
- How deep learning extracts and learns leaf features for plant classification575
This paper learns useful leaf features directly from the raw representations of input data using Convolutional Neural Networks (CNN), and gains intuition of the chosen features based on a Deconvolutional Network (DN) approach, and gains insights into the design of new hybrid feature extraction models which are able to further improve the discriminative power of plant classification systems.
- Total-Text: A Comprehensive Dataset for Scene Text Detection and Recognition571
To evaluate its robustness against curved text, DeconvNet is fine-tuned and benchmarked on Total-Text to facilitate a new research direction for the scene text community.
- A robust arbitrary text detection system for natural scene images562
A robust system based on the concepts of Mutual Direction Symmetry (MDS), Mutual Magnitude Symmetric (MMS) and Gradient Vector Symmeter (GVS) properties to identify text pixel candidates regardless of any orientations including curves from natural scene images is presented.
- Deep-plant: Plant identification with convolutional neural networks487
Experimental results using these CNN features with different classifiers show consistency and superiority compared to the state-of-the art solutions which rely on hand-crafted features.
- Automated Detection and Classification of Oral Lesions Using Deep Learning for Early Detection of Oral Cancer372
Two deep learning based computer vision approaches were assessed for the automated detection and classification of oral lesions for the early detection of oral cancer, these were image classification with ResNet-101 and object detection with the Faster R-CNN.
- 308
- ICDAR2019 Robust Reading Challenge on Arbitrary-Shaped Text - RRC-ArT272
This paper reports the ICDAR2019 Robust Reading Challenge on Arbitrary-Shaped Text - RRC-ArT that consists of three major challenges: i) scene text detection, ii) sceneText recognition, and iii) scenetext spotting.
- 254
- ICDAR 2019 Competition on Large-Scale Street View Text with Partial Labeling - RRC-LSVT207
This competition aims to explore the abilities of state-of-the-art methods to detect and recognize text instances from large-scale street view images, closing gaps between research benchmarks and real applications.
- 193
- A Fusion Approach for Efficient Human Skin Detection188
An eye detector is used to refine the skin model for a specific person and a smoothed 2-D histogram and Gaussian model is combined, for automatic human skin detection in color image(s).
- 177
- Total-Text: toward orientation robustness in scene text detection154
A new scene text detection model is proposed as the baseline for Total-Text, namely Polygon-Faster-RCNN, and its ability to detect text of all orientations is demonstrated.
- Crowd behavior analysis: A review where physics meets biology152
A review of the state-of-the-art methods in crowd behavior analysis from the physics and biologically inspired perspectives is provided, providing insights and comprehensive discussions for a broader understanding of the underlying prospect of blending physics and biology studies in computer vision.
- Deep Polarized Network for Supervised Learning of Accurate Binary Hashing Codes148
A novel deep polarized network (DPN) for learning to hash is proposed, in which each channel in the network outputs is pushed far away from zero by employing a differentiable bit-wise hinge-like loss which is dubbed as polarization loss.
- 146
- On the General Value of Evidence, and Bilingual Scene-Text Visual Question Answering132
A dataset that takes a step towards addressing the VQA problem in that it contains questions expressed in two languages, and an evaluation process that co-opts a well understood image-based metric to reflect the method’s ability to reason.
- ArtGAN: Artwork synthesis with conditional categorical GANs130
The proposed ArtGAN is capable to create realistic artwork, as well as generate compelling real world images that globally look natural with clear shape on CIFAR-10.
- ArtGAN: Artwork Synthesis with Conditional Categorial GANs.127
The proposed ArtGAN is capable to create realistic artwork, as well as generate compelling real world images that globally look natural with clear shape on CIFAR-10.
- 118
- Multi-Organ Plant Classification Based on Convolutional and Recurrent Neural Networks117
A hybrid generic-organ convolutional neural network (HGO-CNN) is introduced, which takes into account both organ and generic information, combining them using a new feature fusion scheme for species classification.
- 101
- 94
- Fuzzy human motion analysis: A review92
The fuzzy set oriented approaches for HMA are reviewed, individuating how the fuzzy set may improve the HMA, envisaging and delineating the future perspectives and ideas for future research are discussed.
- Protecting Intellectual Property of Generative Adversarial Networks from Ambiguity Attacks88
A complete protection framework in both black-box and white-box settings to enforce IPR protection on GANs is presented and it is shown that the proposed method does not compromise the original GAns performance, and at the same time, it is able to withstand both removal and ambiguity attacks against embedded watermarks.
- Deep learning radiomics in breast cancer with different modalities: Overview and future78
A survey on deep learning architectures exploited in breast cancer, including discriminative architectures and generative architectures is provided and some potential challenges along with future research directions are proposed as references to the clinical treatment management and decision making utilizing such breast cancer CAD systems.
- Phrase-based image caption generator with hierarchical LSTM network78
This paper proposes a phrase-based image captioning model using a hierarchical Long Short-Term Memory (phi-LSTM) architecture to generate image description and shows a better or competitive result on the Flickr8k, Flickr30k and MS-COCO datasets in comparison to the state-of the art models.
- Federated Learning with Privacy-preserving and Model IP-right-protection76
FedIPR, a novel ownership verification scheme, is introduced by embedding watermarks into FL models to verify the ownership of FL models and protect model intellectual property rights (IPR or IP-right for short).
- 74
- Rethinking Privacy Preserving Deep Learning: How to Evaluate and Thwart Privacy Attacks73
This paper investigates capabilities of Privacy-Preserving Deep Learning mechanisms against various forms of privacy attacks, and proposes to quantitatively measure the trade-off between model accuracy and privacy losses incurred by reconstruction, tracing and membership attacks.
- Numerical solutions of fuzzy differential equations by an efficient Runge–Kutta method with generalized differentiability71
An extended fourth-order Runge–Kutta method is studied to approximate the solutions of first-order fuzzy differential equations using a generalized characterization theorem and its error analysis guarantees pointwise convergence.
- 65
- Ten Challenging Problems in Federated Foundation Models64
This paper provides a comprehensive summary of the ten challenging problems inherent in FedFMs, encompassing foundational theory, utilization of private data, utilization of private data, continual learning, unlearning, Non-IID and graph data, bidirectional knowledge transfer, incentive mechanism design, game mechanism design, model watermarking, and efficiency.
- A novel multimodal communication framework using robot partner for aging population62
Experimental results have shown the effectiveness of the proposed communication framework in establishing natural communication between the human and the robot partner.
- iSurveillance: Intelligent framework for multiple events detection in surveillance videos60
An intelligent framework for detection of multiple events in surveillance videos is presented, based on the principle of compositionality, which modularize the surveillance problems into a set of variables comprising regions-of-interest, classes, attributes and rules to construct a knowledge-based understanding of the environment.
- Low-light image enhancement using Gaussian Process for features retrieval54
This work proposes to model low-light enhancement as a set of localized functions using Gaussian Process that is trained at runtime using data from a simple Convolutional Neural Network to provide the necessary feature information as reference, and proposes two new evaluation metrics to better assess enhancement algorithms to support high level computer vision tasks.
- Protect, show, attend and tell: Empowering image captioning models with ownership protection51
This work is the first to propose ownership protection on image captioning task and studies and proposes two different embedding schemes in the hidden memory state of a recurrent neural network to protect the image Captioning model.
- A novel character segmentation-reconstruction approach for license plate recognition48
Experimental results on benchmark license plate databases, as well as video databases, namely, ICDAR 2015, YVT video, and natural scene data, show that the proposed technique is effective and useful.
- Crowd Saliency Detection via Global Similarity Structure45
A novel framework to identify and localize salient regions in a crowd scene, by transforming low-level features extracted from crowd motion field into a global similarity structure representation, which allows the discovery of the intrinsic manifold of the motion dynamics, which could not be captured by the low- level representation.
- A weighted inference engine based on interval-valued fuzzy relational theory43
A fuzzification method that able to fuzzify the input data and also train the inference engines is developed, so the BK subproduct-based inference systems can be built without human intervention, which are cumbersome and time consuming.
- Successive approximation method for Caputo q-fractional IVPs42
This paper proposes q-Krasnoselskii–Krein type condition to investigate the convergence of the Caputo q-fractional derivatives method and suggests that under which conditions this method will be convergence.
- Non-fragile state observer design for neural networks with Markovian jumping parameters and time-delays42
A delay-dependent stability criterion is obtained in terms of linear matrix inequalities (LMIs) by developing a reciprocal convex approach and based on the Lyapunov–Krasovskii functional, and stochastic stability theory.
- Design of efficient blue phosphorescent bottom emitting light emitting diodes by machine learning approach41
It is shown that the random forest algorithm (a machine learning approach) provides significant improved predictive performance over linear regression analysis and other multivariate regression models such as extreme gradient boosting, adaptive boosting, gradient boosting and k-nearest neighbor.
- Recognition of Human Motion From Qualitative Normalised Templates40
A Qualitative Normalised Templates framework for solving the human motion classification problem, which replaces the matching algorithm with the proposed QNTs to categorise complex human motion into sets of fuzzy qualitative angles and positions in quantity space.
- Script independent approach for multi-oriented text detection in scene image37
Experimental results show that the proposed RRT for text detection in multi-oriented and multi-script environments outperforms the existing methods in terms of multi-lingual andMulti-oriented text detection ability.
- Plant Identification System based on a Convolutional Neural Network for the LifeClef 2016 Plant Classification Task.37
The architecture of the plant classification system that uses a convolutional neural network for the LifeClef 2016 challenge, which is to identify 1000 species of images of plants corresponding to 7 different plant organs, is described.
- MTSN: A Multi-Temporal Stream Network for Spotting Facial Macro- and Micro-Expression with Hard and Soft Pseudo-labels34
It is demonstrated how a single output from the MTSN model can be post-processed to predict both macro- and micro-expression intervals, and a hard and soft pseudo-labeling technique is introduced to enable the network to distinguish expression frames from non-expression frames via the learning of salient features in the expression peak frame.
- phi-LSTM: A Phrase-Based Hierarchical LSTM Model for Image Captioning33
This paper proposes a phrase-based hierarchical Long Short-Term Memory model to generate image description, adopting a convolutional neural network to learn image features and the LSTM to learn the word sequence in a sentence.
- 33
- An Automated Patient Self-Monitoring System to Reduce Health Care System Burden During the COVID-19 Pandemic in Malaysia: Development and Implementation Study32
This study demonstrated that developing a COVID-19 symptom monitoring system within a short time during a pandemic is feasible using the agile development method.
- Assessing agreement in measurements of orthodontic study models: Digital caliper on plaster models vs 3-dimensional software on models scanned by structured-light scanner32
Measurements made on models scanned by the 3D structured-light scanner were in good agreement with those made on conventional plaster models and were, therefore, clinically acceptable.
- 32
- Semantic Matching Efficiency of Supply and Demand Texts on Online Technology Trading Platforms: Taking the Electronic Information of Three Platforms as an Example29
OtTPs can help the suppliers and demanders to retrieve information accurately, and assist the OTTPs to carry out trade promotion and evaluate trade performance, according to the proposed Supply-Demand Matching Efficiency (SDME).
- InteractDiffusion: Interaction Control in Text-to-Image Diffusion Models27
A pluggable interaction control model is proposed that extends existing pre-trained T2I diffusion models to enable them being better conditioned on interactions, and outperforms existing baselines by a large margin in HOI detection score, as well as fidelity in FID and KID.
- 26
- ACORT: A compact object relation transformer for parameter efficient image captioning26
Three parameter reduction methods for image captioning Transformers: Radix Encoding, cross-layer parameter sharing, and attention parameter sharing are presented, resulting in proposed ACORT models that have 3.7x to 21.6x fewer parameters than the baseline model without compromising test performance.
- Yuan: Yielding Unblemished Aesthetics Through a Unified Network for Visual Imperfections Removal in Generated Images25
Yuan, a novel framework that autonomously corrects visual imperfections in text-to-image synthesis, demonstrates superior performance in eliminating visual imperfections and has potential to significantly enhance the quality and applicability of AI-generated images across diverse fields.
- Fine-Tuning Deep Learning Architectures for Early Detection of Oral Cancer25
This paper extended on the previous deep learning work and focused on the binary image classification of ‘referral’ vs. ‘non-referral’, with improvements to the previous work achieved.
- 24
- Human Motion Sensing and Recognition24
These techniques are shown to effectively address the above challenges by bridging the gap between symbolic cognitive functions and numerical sensing & control tasks in intelligent systems.
- Utility and usability of an automated COVID-19 symptom monitoring system (CoSMoS) in primary care during COVID-19 pandemic: A qualitative feasibility study23
CoSMoS is feasible and useful to patients and doctors in providing remote monitoring and teleconsultation during the COVID-19 pandemic and the utility and usability evaluation enables the refinement of CoSMoS to be a patient-centred monitoring system.
- End-to-End Supermask Pruning: Learning to Prune Image Captioning Models23
This work provides an extensive comparison of various unstructured weight pruning methods on three different popular image captioning architectures, namely Soft-Attention, Up-Down and Object Relation Transformer, and proposes a novel end-to-end weight pruned method that performs gradual sparsification based on weight sensitivity to the training loss.
- 23
- 21
- MSFNet: A Multi-Scale Space-Time Frequency Fusion Network for Motor Imagery EEG Classification20
A Multi-scale Space-time Frequency fusion Network (MSFNet) is proposed to improve the MI-EEG classification accuracy and realized the extraction and efficient fusion of spatiotemporal and frequency domain features and the high-precision MI-EEG classification, which has an important application value.
- Refined particle swarm intelligence method for abrupt motion tracking20
This work casts tracking as an optimisation problem and proposes a novel abrupt motion tracker based on swarm intelligence – the SwATrack and introduces an optimised swarm-based sampling strategy for a tradeoff between the exploration and exploitation of the state space in search for the optimal proposal distribution.
- Early detection of human actions—A hybrid approach19
This study aims at training a detector to be capable of recognizing a human action when only partial action sample is seen, and proposes a hybrid technique which combines the benefits of computer vision as well as fuzzy set theory based on the fuzzy Bandler and Kohout's sub-triangle product (BK subproduct).
- 18
- Association Rules Mining for Hospital Readmission: A Case Study18
There is a significant correlation between potential readmission factors (threshold of various settings for readmission length) and basic demographic variables and association rule mining (ARM) was utilised to extract the hidden input variable patterns and relationships among admitted patients.
- Action recognition on continuous video18
A novel framework, named as long-term video action recognition (LVAR) to perform generic action classification in the continuous video by introducing a partial recurrence connection to propagate information within every layer of a spatial-temporal network, such as the well-known C3D.
- HGO-CNN: Hybrid generic-organ convolutional neural network for multi-organ plant classification18
This work introduces a hybrid generic-organ convolutional neural network (HGO-CNN), which takes into account both organ and generic information, combining them using a new feature fusion scheme for species classification, and shows that the proposed method outperforms the state-of-the-art results.
- 17
- 16
- Color video denoising using epitome and sparse coding16
This paper presents a unified framework to deal with video denoising problems by adopting a two-steps process, namely the video epitome and sparse coding, and shows comparable results to conventional methods in both spatial and transform domain.
- A Runge–Kutta method with reduced number of function evaluations to solve hybrid fuzzy differential equations16
A numerical algorithm is employed to solve first-order hybrid fuzzy differential equation (HFDE) based on the high order Runge–Kutta method and the characterization theorem of Bede is considered to solve the HFDE numerically.
- 15
- D’OraCa: Deep Learning-Based Classification of Oral Lesions with Mouth Landmark Guidance for Early Detection of Oral Cancer15
Quantitative and qualitative results demon-strate the importance of the mouth landmark detection model in guiding the classification model to classify the oral lesions into four referral decision classes.
- Human Motion Sensing and Recognition: A Fuzzy Qualitative Approach15
These techniques are shown to effectively address the above challenges by bridging the gap between symbolic cognitive functions and numerical sensing & control tasks in intelligent systems.
- OERScout technology framework: A novel approach to open educational resources search15
The objectives of the work are to develop a technology framework which will parametrically measure the usefulness of an OER for a particular academic purpose based on the openness, accessibility, and relevance attributes, and provide academics with a mechanism to locate OER which are of an acceptable academic standard.
- Clinically Guided Trainable Soft Attention for Early Detection of Oral Cancer14
The use of convolutional neural networks with trainable soft attention is proposed, and the use of localization loss to penalize the difference between attention maps and clinically annotated mask to help guide soft attention.
- 14
- Detection of salient regions in crowded scenes14
The experimental results on public, crowd scenes have shown the effectiveness of the proposed framework in detecting salient regions which correspond to an unstable flow, occlusions, bottlenecks, and entries and exits.
- 14
- Image collection and annotation platforms to establish a multi‐source database of oral lesions13
This is the first description of a database with clinically-labelled oral lesions, which could accelerate the improvement of AI algorithms that can promote the early detection of high-risk oral lesions.
- Relating molecular descriptors to frontier orbital energy levels, singlet and triplet excited states of fused tricyclics using machine learning13
This research demonstrated that machine learning and data analytics in guiding how certain chemical substructures correlate with the molecule energy states can help to improve the material performance.
- 13
- Fuzzy qualitative deep compression network13
This paper proposes a novel one-shot deep compression method based on the fuzzy quantity space to remove redundant CNN weights and presents the first attempt to train an end-to-end fuzzy qualitative deep compression model in the fine-art paintings classification problem.
- Robust synchronization of uncertain chaotic neural networks with randomly occurring uncertainties and non-fragile output coupling delayed feedback controllers13
An improved Lyapunov–Krasovskii functional is introduced and a delay-dependent non-fragile output feedback controller is designed to achieve synchronization with the help of a drive–response system and the linear matrix inequality approach.
- 13
- Extremely Low-Light Image Enhancement with Scene Text Restoration12
A novel image enhancement framework is proposed to precisely restore the scene texts, as well as the overall quality of the image simultaneously under extremely low-light conditions, using a self-regularised attention map, an edge map, and a novel text detection loss.
- Hospital readmission prediction based on improved feature selection using grey relational analysis and LASSO12
A robust hospital readmission prediction framework by combining the feature selection algorithm and machine learning (ML) classifiers and the improved feature selection is proposed by considering the uncertainty in patient's attributes that leads to the output variable.
- 12
- 12
- 11
- User acceptance of a touchless sterile system to control virtual orthodontic study models11
An evaluation of user acceptance of a hand-gesture-based touchless sterile system for interaction with and control of a set of 3-dimensional digitized orthodontic study models using the Kinect motion-capture sensor supports the adoption of this promising development for a sterile touch-free patient record-management system.
- Colour Object Classification Using the Fusion of Visible and Near-Infrared Spectra11
A Fuzzy-Genetic colour processing algorithm that adaptively marries together the visible and near-infrared spectra signals for the purpose of colour object recognition yields better colour classification results than utilizing the pure visible spectrum alone.
- 10
- 9
- Unprecedented Usage of Pre-trained CNNs on Beauty Product9
This paper adopts several well known deep learning models pre-trained on ImageNet, and evaluates their performance using different distance metrics to secure fourth position in the Perfect Corp Challenge.
- LifeClef 2017 Plant Identification Challenge: Classifying Plants using Generic-Organ Correlation Features.9
The objective of the challenge is to evaluate to what extent machine learning and computer vision can learn from noisy data compared to trusted data, and the recent proposed hybrid generic-organ convolutional neural network, abbreviated HGO-CNN is employed to train on different composition of plant datasets.
- SwATrack: A Swarm Intelligence-based Abrupt Motion Tracker9
This work proposes a novel abrupt motion tracker that is based on swarm intelligence - the SwATrack and introduces an optimised swarm-based sampling strategy to enrich the trade-off between the exploration and exploitation of the search space in search for the optimal proposal distribution.
- RECENT ADVANCES IN FUZZY QUALITATIVE REASONING9
An eye detector is used to refine the skin model for a specific person and a smoothed 2-D histogram and Gaussian model is combined, for automatic human skin detection in color image(s).
- Logical Connectives and Operativeness of BK Sub-triangle Product in Fuzzy Inferencing9
This paper examines this sub-triangle product in a medical expert system where lists of equations were initialized from the sub-Triangle product as the inference structures, and proposed a combination of inference structures that good for forming inference engines of medical expert systems.
- 8
- 8
- Recognizing unknown objects with attributes relationship model8
This paper combines both the strength of attributes and image hierarchy by proposing Attributes Relationship Model (ARM) to perform zero-shot learning and achieves state-of-the-art accuracy in AwA dataset.
- 8
- 7
- Plant Classification Based on Gated Recurrent Unit7
A new framework for plant structural learning using the recurrent neural network (RNN) approach, which supports classification based on a varying number of plant views composed of one or more organs of a plant, by optimizing the dependencies between them.
- Granular-based dense crowd density estimation7
This paper proposes a novel approach to learn discriminative crowd features from granules, that conforms to the outline between crowd and background (i.e. scene independent), and shows that by studying the inner statistics of granules for density estimation, this approach is adaptive to arbitrary distribution of crowd.
- 7
- 7
- 7
- A Real-Time Vision-Based Framework for Human-Robot Interaction7
A real-time vision-based framework to detect human face and analysis of the human face direction in window area to interact with robot and an intelligent wheelchair whose motion can be controlled by the user's face direction is presented.
- 6
- 6
- Coherent group detection in still image6
A novel framework that merges these two classical approaches to achieve a higher level of crowd understanding, i.e., detecting coherent groups within a crowd in a still image, can infer crowd segments at both image-level and coherent group level.
- Orthogonal planar search (OPS) for coronary artery centerline extraction6
In this paper, orthogonal planar search for coronary artery centerline extraction is investigated to assist in coronary artery diseases diagnosis and the results quantitatively and qualitatively are demonstrated.
- 6
- 6
- Human Arm-Motion Classification Using Qualitative Normalised Templates6
Experimental evaluation demonstrates that the proposed method is effective for the classification of human-arm motion and future work will focus on extending the proposed approach to classify human arm motion using qualitative normalized templates.
- Enhancing Facial Expression Recognition Under Data Uncertainty Based on Embedding Proximity5
A new framework based on embedding proximity is proposed to ensure consistent updating of feature representations with rectifications made during relabeling to overcome this limitation and demonstrate significant improvements in addressing inconsistency issues and enhancing overall performance.
- 5
- Optimizing Workflow Task Clustering Using Reinforcement Learning5
This paper proposes the use of reinforcement learning to tackle this problem by automating the discovery of the optimal clustering number for the submitted workflow by model the workflow environment that allows the reinforcement learning agent to interact by determining the cluster number for every round of workflow execution.
- 5
- Fuzzy Empirical Copula for Estimating Data Dependence Structure5
This chapter proposes a new version of Fuzzy Clustering by Local Approximation of Memberships (FLAME) which is proposed to be integrated into empirical copula to reduce the computation time of dependence structure estimation.
- 5
- Enhanced Random Forest with Image/Patch-Level Learning for Image Understanding5
This paper proposes a novel way to update the RF codebook learning for a more discriminative codebook with the introduction of the soft class labels, estimated from the pLSA model based on a feedback scheme.
- Banking Done Right: Redefining Retail Banking with Language-Centric AI4
Ryt AI is presented, an LLM-native agentic framework that powers Ryt Bank to enable customers to execute core financial transactions through natural language conversation, demonstrating that regulator-approved natural-language interfaces can reliably support core financial operations under strict governance.
- A Survey of Challenges and Opportunities in Vertical Federated Unlearning4
The first comprehensive survey of VFU methods is presented, identifying key challenges in the field, including high resource demands, limited support for complex unlearning tasks, and inconsistent benchmarks.
- Radioport: a radiomics-reporting network for interpretable deep learning in BI-RADS classification of mammographic calcification4
A novel model called radiomics-reporting network (Radioport), which incorporates text attention to improve the interpretability of DLR in mammographic calcification diagnosis, and can explicitly simulate the mammographic calcification diagnosis process.
- 4
- ICDAR 2021 Competition on Integrated Circuit Text Spotting and Aesthetic Assessment4
A text on chips dataset, ICText is used as the main target for the proposed Robust Reading Challenge on Integrated Circuit Text Spotting and Aesthetic Assessment (RRC-ICText) 2021 to encourage the research on this problem.
- 4
- 4
- Fuzzy qualitative human model for viewpoint identification4
Experimental results showed the reliability and effectiveness of the proposed viewpoint estimation framework by using the challenging IXMAS human action dataset and a novelhuman contour descriptor namely the fuzzy qualitative human contour is extracted from the fuzzy quantitative Poisson human model for viewpoint analysis.
- Keybook: Unbias object recognition using keywords4
An improved codebook representation in the Bag-of-Words (BoW) approach by generating Keybook, which is composed from the keywords that significantly represent the object classes and will not bias to any of the domain.
- 4
- 4
- A Comparison Of Posture Recognition Using Supervised And Unsupervised Learning Algorithms4
Experimental results have shown that the supervised learning algorithms outperform the unsupervised learning algorithms in terms of the number of correctly classified postures and will focus on detecting abnormal behaviour based on these recognised static postures.
- 4
- 4
- 3
- IPR-NeRF: Ownership Verification meets Neural Radiance Field3
This paper proposes a comprehensive intellectual property protection framework for the NeRF model in both black-box and white-box settings, namely IPR-NeRF, and demonstrates that not only does the approach maintain the fidelity of IPR-NeRF models, but it is also robust against both ambiguity and removal attacks compared to prior arts.
- 3
- 3
- 3
- The establishment of a multi-source dataset of images on common oral lesions3
This work describes the establishment of a multi-sourced image dataset through the development of a platform for the collection and annotation of images and describes the first description of a database with well-annotated oral lesions.
- Regularization-based modulation transfer function compensation for optical satellite image restoration using joint statistical model in curvelet domain3
This work exploits the merit of image prior characteristics in both local smoothness and nonlocal self-similarity properties of an image in a hybrid domain to design effective regularization terms that reflect these image properties.
- 3
- 3
- 3
- On-orbit spatial characterization based on satellite image structure3
A segmentation method to select the ideal candidates for MTF estimation, an adaptive structure selection method that removes detrimental structures and selects useful information for point spread function (PSF) estimation, and a robust estimation method that can warrant a reliable result for on-orbit spatial characterization.
- 3
- Recognizing Constrained 3D Human Motion: An Inference Approach3
This chapter presents a novel inference engine for recognising occluded 3D human motion assisted by the recognition context via a novel method called fuzzy quantile generation which employs metrics derived from the probabilistic quantile function.
- 3
- 3
- Fuzzy-rough based decision system for gait adopting instance selection3
The aim is to adopt instance selection methods to eliminate redundant or noisy sample from training data thus preventing the system from detecting erroneous gait patterns or deviations, and to improve the efficacy of various classifiers in detecting distracted gait.
- An improved BK sub-triangle product approach for scene classification3
An improved Bandler and Kohout’s sub-triangle product (BK subproduct) approach to address scene classification, which is able to model non-mutually exclusive data, as opposed to conventional solutions.
- 3
- Computational Intelligence for Human Interactive Communication of Robot Partners3
An integration method for estimating human behaviors using sound source angle information, and gesture recognition by the multi-layered spiking neural network with the time series of human hand positions is proposed.
- 3
- Gorgeous: Creating narrative-driven makeup ideas via image prompts2
Gorgeous is the first to empower users to incorporate narrative elements into makeup ideation via image-based prompts, allowing for the creation of makeup concepts that visually convey user expressions, offering innovative and imaginative makeup ideas for real-world application.
- 2
- Unsupervised Hashing with Similarity Distribution Calibration2
A novel Similarity Distribution Calibration (SDC) method is introduced, which aligns the hash code similarity distribution towards a calibration distribution with sufficient spread across the entire similarity capacity/range, to alleviate the similarity collapse problem.
- 2
- Predictive modelling of hospital readmission: Evaluation of different preprocessing techniques on machine learning classifiers2
This study aims to analyse the impact of different preprocessing methods on the performance of different machine learning classifiers and indicated that the preprocessing approaches helped improve the model’s prediction of hospital readmission.
- 2
- 2
- Protect, Show, Attend and Tell: Image Captioning Model with Ownership Protection.2
It is demonstrated that current digital watermarking framework is insufficient to protect image captioning task that often regarded as one of the frontier A.I. problems, and a proposed embedding schemes in the hidden memory state of a recurrent neural network to protect images captioning model is the first to propose ownership protection.
- Learning a Generative Adversarial Network for High Resolution Artwork Synthesis.2
A series of new approaches to improve Generative Adversarial Network (GAN) for conditional image synthesis are proposed and the proposed model is named as ArtGAN, which is able to generate plausible-looking images on Oxford-102 and CUB-200 and able to draw realistic artworks based on style, artist, and genre.
- Fuzzy Qualitative Trigonometry2
In order to bridge the gap between qualitative and quantitative descriptions of physical systems, a fuzzy qualitative representation of trigonometry (FQT) is proposed, which provides theoretical foundations for the representationof trigonometric properties.
- 2
- 2
- 2
- 2
- 2
- MeMoSA dataset: A multi-country collection of over 30,000 oral mucosa images with clinically labelled lesions1
The MeMoSA dataset will serve as a significant resource to drive the training, evaluation, and refinement of AI-driven diagnostic algorithms, potentially improving diagnostic accuracy and enabling rigorous benchmarking against clinical expert assessments, for the early detection of oral cancer.
- Reference-free human-object interaction editing1
InteractEdit is presented, a novel framework for reference-free Human-Object Interaction (HOI) editing that tackles the challenging task of transforming an existing interaction in an image into a new, desired interaction while preserving the identities of the subject and object.
- Single-Pass, Depth-Selective Reading for Multi-Aspect Sentiment Analysis1
It is argued that Transformer depth is a costly, queryable resource, and DABS, a single-pass inference framework that encodes each sentence once to construct a reusable, depth-ordered substrate, is proposed, which decouples shared sentence encoding from lightweight, aspect-conditioned readout.
- VAGNet: Grounding 3D Affordance from Human-Object Interactions in Videos1
VAGNet is proposed, a framework that aligns video-derived interaction cues with 3D structure to resolve ambiguities that static cues cannot address, and PVAD is introduced, the first HOI video-3D pairing affordance dataset, providing functional supervision unavailable in prior works.
- They Are Not the Same: Direct Causes Are Not Grounded Emotion Explanations1
Binary-trained models assign higher pair scores to nearby lexically similar non-pair candidates than to evidence supported but structurally harder emo-cause and emo-context pairs, which can reward convenient attributions over grounded explanations.
- 1
- Gorgeous: Create Your Desired Character Facial Makeup from Any Ideas1
Gorgeous$, a novel diffusion-based makeup application method that goes beyond simple transfer by innovatively crafting unique and thematic facial makeup, opens up new possibilities for integrating broader story elements into character makeup, thereby enhancing the narrative depth and visual impact in storytelling.
- Everyone Can Attack: Repurpose Lossy Compression as a Natural Backdoor Attack1
This paper shows that there exists a more severe backdoor threat: anyone can exploit an easily-accessible algorithm for silent backdoor attacks, and can employ the widely-used lossy image compression from a plethora of compression tools to effortlessly inject a trigger pattern into an image without leaving any noticeable trace.
- Synthesizing Physical Backdoor Datasets: An Automated Framework Leveraging Deep Generative Models1
A recipe that empowers backdoor researchers to effortlessly create a malicious, physical backdoor dataset based on advances in generative modeling is unleashed, which effectively mitigates the perceived complexity associated with creating a physical backdoor datasets, transforming it from a daunting task into an attainable objective.
- 1
- 1
- 1
- 1
- 1
- OERScout: widening access to OER through faceted search1
This paper discusses how the OERScout technology framework uses a “faceted search” approach to locate the most desirable OER from sources spread throughout the globe and highlights how focused searching can greatly improve access to OER readily useable in teaching and learning.
- A Neighbourhood Search Feedback for Coronary Artery Centerline Tracking1
A neighbourhood search feedback mechanism in centerline tracking to overcome error propagation and seed point optimization where the defined seed point is optimized before tracking is initialized and Backward-Forward Correlation Verification to verify the points before propagate in the tracking process.
- OERScout: Autonomous Clustering of Open Educational Resources1
The design and development of a clustering algorithm which accurately clusters text based OER materials by building a Keyword-Document Matrix (KDM) using autonomously identified domain specific keywords is discussed.
- Hybrid particle swarm optimisation for data clustering1
This paper explores one such approach which is hardly known outside the search heuristic field - the Particle Swarm Optimisation+k-means (PSOk) for this purpose and demonstrates the effectiveness of the proposed solution in terms of precision and computational complexity.
- Anomaly Detection over Spatiotemporal Object Using Adaptive Piecewise Model1
This paper presents a novel anomaly detection framework to detect anomalous motion trajectory using the fusion of adaptive piecewise analysis and fuzzy rule-based method, which has validated the effectiveness of the proposed method.
- Adapting robot kinematics for human-arm motion recognition1
A novel method to the analysis of human-arm motion, in particular improving the efficiency of conventional motion recognition algorithms, where qualitative normalised templates (QNTs) is proposed to replace the conventional approaches.
- Hidden Consensus:Preference-Validity Compression in Human Feedback–
It is argued that future alignment methods should satisfy Validity-Preserving Consistency, remaining stable across plural-valid interpretive frames rather than collapsing them into a single reward target.
- –
- OneHOI: Unifying Human-Object Interaction Generation and Editing–
OneHOI is introduced, a unified diffusion transformer framework that consolidates HOI generation and editing into a single conditional denoising process driven by shared structured interaction representations, achieving state-of-the-art results across both HOI generation and editing.
- –
- –
- Rethinking Molecular Graph Backdoors under Chemistry-aware Admission–
The proposed ChemBack is an admission-aware molecular backdoor attack that constructs chemically feasible motif-anchor attachments and ranks admitted candidates by fingerprint-based Tanimoto similarity to clean target-class molecules, and achieves high attack success with fully admitted poisons while preserving clean accuracy.
- Erased, but Not Gone: Output Forgetting Is Not True Forgetting–
The theoretical analysis and empirical results show that standard output-level evaluation can systematically overestimate the success of unlearning, and suggest that current MU is often evaluated for apparent forgetting rather than retraining-consistent forgetting.
- –
- –
- –
- –
- –
- Protégé: Learn and Generate Basic Makeup Styles with Generative Adversarial Networks (GANs)–
Prot e, a new makeup application, leveraging recent generative model -- GANs to learn and automatically generate makeup styles, is proposed, marking a significant leap in digital makeup technology.
- –
- –
- –
- –
- –
- –
- Special issue on “Green and Human Information Technology 2019”–
Five publications that represent the state-of-the-art research ranging from theory to practice are selected, which include a new approach to solve hard exudate detection which will in turn solve the abnormality detection in diabetic retinopathy.
- –
- –
- –
- Feature Extraction for the Identification of Two-Class Mechanical Stability Test of Natural Rubber Latex–
A new feature set for a computer vision-based mechanical stability classification system that is based on the current standard test and demonstrated that the proposed system was able to provide good classification accuracies for this two-class MST problem.
- –
- –
- Restoration of Blurred-Noisy Images Through the Concept of Bilevel Programming–
Numerical measurements of the image quality reveal that the algorithm is comparable with other state-of-the-art methods and has the advantage for image contrast and preserving edge details.
- Analytical Design and Applications of Double-Directional IIR Filters.–
A neighbourhood search feedback mechanism in centerline tracking to overcome error propagation and Seed Point Optimization where the defined seed point is optimized before tracking is initialized and Backward-Forward Correlation Verification to verify the points before propagate in the tracking process.
- Articulated Human Motion Tracking with Online Appearance Learning–
This work employs particle swarm optimization framework to avoid the need of motion model and proposes a novel appearance learning strategy to learn the appearance of each body part in real time and also proposes an appearance model to represent the shape of eachBody part.
- –
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.