Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks89 from public data
- Development and Validation of a Deep Learning System to Detect Glaucomatous Optic Neuropathy Using Fundus Photographs304
Application of GD-CNN to fundus images from different settings and varying image quality demonstrated a high sensitivity, specificity, and generalizability for detecting GON, suggesting that automated DLS could enhance current screening programs in a cost-effective and time-efficient manner.
- 303
- Attention Based Glaucoma Detection: A Large-Scale Database and CNN Model274
The proposed AG-CNN approach significantly advances state-of-the-art glaucoma detection, including an attention prediction subnet, a pathological area localization subnet and a glauca classification subnet.
- 267
- DeepVS: A Deep Learning Based Video Saliency Prediction Approach191
A large-scale eye-tracking database of videos (LEDOV), which includes 32 subjects’ fixations on 538 videos, is established and an object-to-motion convolutional neural network (OM-CNN) is developed to predict the intra-frame saliency for DeepVS, which is composed of the objectness and motion subnets.
- Joint Learning of 3D Lesion Segmentation and Classification for Explainable COVID-19 Diagnosis78
Experimental results show that the joint learning framework in the DeepSC-COVID model significantly improves the performance of 3D lesion segmentation and disease classification in both efficiency and efficacy.
- 78
- Joint Learning of Multi-Level Tasks for Diabetic Retinopathy Grading on Low-Resolution Fundus Images73
- 73
- 67
- Deep Multi-Task Learning for Diabetic Retinopathy Grading in Fundus Images55
A deep multi-task learning based DR grading (DeepMT-DR) method for LR fundus images, which simultaneously handles the auxiliary tasks of ISR and lesion segmentation and the high-level task of DR grading.
- Uncertainty Guided Adaptive Warping for Robust and Efficient Stereo Matching54
A novel Uncertainty Guided Adaptive Correlation (UGAC) module is introduced to robustly adapt the same model for different scenarios and a variance-based uncertainty estimation is employed to adaptively adjust the sampling area during warping operation.
- 47
- 45
- 42
- NTIRE 2022 Challenge on Super-Resolution and Quality Enhancement of Compressed Video: Dataset, Methods and Results41
This paper reviews the NTIRE 2022 Challenge on Super-Resolution and Quality Enhancement of Compressed Video and proposed the LDV 2.0 dataset, which includes theLDV dataset (240 videos) and 95 additional videos.
- 40
- 38
- 34
- Image Saliency Prediction in Transformed Domain: A Deep Complex Neural Network Method25
A novel deep complex neural network, named SalDCNN, to predict image saliency by learning features in both pixel and transformed domains, including two main stages: the complex dense encoder and three-stream multi-domain decoder.
- 24
- A multimodal transformer system for noninvasive diabetic nephropathy diagnosis via retinal imaging23
A transformer-based deep learning system for detecting DN and NDRD upon non-invasive multi-modal data of fundus images and clinical characteristics is developed and the nephrologists’ diagnosis accuracy can be improved by 21%, through visualization assistance of the DL system.
- Bottom-up saliency detection with sparse representation of learnt texture atoms22
It is validated that the proposed SR-LTA feature for bottom-up saliency detection, in light of the learnt salient and non-salient dictionaries, can advance state-of-the-art saliency Detection on natural images.
- DeepVS2.0: A Saliency-Structured Deep Learning Method for Predicting Dynamic Visual Attention20
A novel DNN-based video saliency prediction method, called DeepVS2.0, is proposed, which establishes a large-scale eye-tracking database of videos (LEDOV) and develops a saliency-structured convolutional long short-term memory network (SS-ConvLSTM), to consider the center-bias and sparsity of human attention maps.
- Progressive Training of A Two-Stage Framework for Video Restoration16
This work proposed a two-stage framework including a multi-frame recurrent network and a single-frame transformer to cope with the training of recurrent neural networks and transformers, which wins two champions and a runner-up in the NTIRE 2022 super-resolution and quality enhancement of compressed video challenges.
- 15
- 15
- Multi-Task Learning-Based Immunofluorescence Classification of Kidney Disease14
The novel MTL method not only diagnosed four types of kidney diseases through blurred IF images but also showed good performance in two auxiliary tasks: image quality assessment and de-blurring.
- 12
- 11
- 10
- 10
- 10
- DeepCT: A novel deep complex-valued network with learnable transform for video saliency prediction10
This paper develops a novel deep Complex-valued network with learnable Transform (DeepCT) for video saliency prediction, which outperforms other 13 state-of-the-art methods for videoSaliency prediction.
- A Strategy of Particle-Pair for Vector Quantization in Image Coding10
This strategy prevents the particle from being trapped in a local optimum, memoryizes and estimates the best direction the particle moves toward to find the optimum codebook design, and the dependence of the final optimum codebooks on the selection of the initial codebook is reduced effectively.
- Deep multi-task learning for nephropathy diagnosis on immunofluorescence images8
The proposed hierarchical multi-task learning framework provides the new scope to narrow the gap between the low-level vision and high-level medical tasks, and will contribute to nephropathy diagnosis in clinical scenarios.
- 7
- Multi-modal transcriptomics: integrating machine learning and convolutional neural networks to identify immune biomarkers in atherosclerosis7
This study introduces three signature genes for atherosclerosis, shaping a novel paradigm for investigating clinical immunological medications, and distinguishes the high biocidal C2 subtype from the inflammation-modulating C1 subtype, utilizing identified signature gene as crucial targets.
- 7
- 6
- 6
- 6
- 5
- Lightweight wood panel defect detection method incorporating attention mechanism and feature fusion network5
A lightweight wood panel defect detection method called YOLOv5-LW, which incorporates attention mechanisms and a feature fusion network, and enhances the backbone network by incorporating Efficient Channel Attention (ECA), which improves the network's focus on key information relevant to defect detection.
- 5
- 4
- Patient-specific dynamic reference frame for navigation-assisted surgery in mandible: a novel noninvasive technical method4
The PS-DRF system offers a convenient, effective, and adaptable approach by integrating navigation technology with 3D printing that has the potential to simplify navigation-assisted mandibular surgery and facilitate the broader clinical implementation of computer-assisted navigation in maxillofacial procedures.
- A multi-task deep sequential neural network for IgA nephropathy Oxford classification and prognosis prediction4
DeepSNN successfully integrates multiple diagnostic tasks with performance comparable to senior pathologists, demonstrating substantial potential for streamlining IgAN clinical workflows and addressing critical gaps in automated renal pathology analysis while maintaining clinical interpretability.
- MT-VQA: A Multi-task Approach for Quality Assessment of Short-form Videos4
A two-branch multi-task VQA approach, MT-VQA, is proposed, in which both tasks of VQA and video saliency prediction (VSP) can be accomplished for short-form videos.
- 4
- 4
- 4
- An improved second-generation digital image watermarking scheme4
An improved second generation digital image watermarking scheme that exploits the region feature instead of point or line feature and is robust against compression, noise intrinsically and more robust against geometrical attacks and JPEG compression compared with Kutter's method.
- 3
- Hierarchical Semantic Compression for Consistent Image Semantic Restoration3
This work proposes a novel hierarchical semantic compression (HSC) framework that purely operates within intrinsic semantic spaces from generative models, which is able to achieve efficient compression for consistent semantic restoration in image/video compression paradigms.
- 3
- 3
- A FPGA-Based Real Time QRS Complex Detection System Using Adaptive Lifting Scheme3
A real time QRS wave detection system implemented using the FPGA based ALS system based on the Adaptive Lifting Scheme, the digitized ECG sequence is directly processed in the spatial domain, thus greatly simplifies the system design and reduces the memory usage.
- 2
- An improved second generation digital image watermarking scheme based on watershed transform2
An improved second generation digital image watermarking scheme is proposed, which exploits the region feature instead of point or line feature and is robust against compression, noise intrinsically and more robust against geometrical attacks and JPEG compression.
- The Illusion of Forgetting: Attack Unlearned Diffusion via Initial Latent Variable Optimization1
IVO (Initial Latent Variable Optimization), a novel attack framework designed to assess the robustness of current unlearning methods, is proposed, which optimizes initial latent variables to realign the noise distribution of unlearned models with that of their vanilla counterparts, which reconstructs the fractured mappings and consequently revives dormant memories.
- 1
- 1
- 1
- Microstructure and Properties of Ni-TiCN Composite Coatings on Ti-6Al-4V Substrate by Laser Cladding1
- 1
- 1
- Research on Medical Image Segmentation Method Based on Improved U-Net3+1
An improved full scale skip connection network structure for the CT liver image segmentation task is proposed, which includes a biomimetic attention module between the shallow encoder and the deep decoder and the feature fusion proportion coefficient between the two is learned to enhance the attention of the overall network to the segmented target area.
- 1
- 1
- 1
- Combined Use of Spectral and Spatial Features for Building Extraction in Multi-Spectral Imagery1
An advanced algorithm based on the combined use of spectral and spatial features will be developed and employed to recognize and extract buildings from multi-spectral imagery in this paper.
- A multiple digital watermarking algorithm based on 1D and 2D chaotic sequences1
A multiple watermarking algorithm based on 1-D and 2-D chaotic sequences is proposed that is robust to many common attacks and it is a reliable copyright protection for multiple legal owners.
- –
- –
- –
- –
- Research on Polarized Image Fusion Algorithm Based on Image Enhancement–
Experimental results demonstrate that the fused images produced by this algorithm not only achieve optimal values across multiple objective evaluation metrics but also exhibit superior visual quality that aligns better with human visual perception.
- –
- –
- Noninvasive Extraction of Maternal and Fetal Electrocardiograms Using Progressive Periodic Source Peel-off–
A periodic constrained FastICA(pcFastICA), developed to improve the precision of examining and correcting FECG source signals, based on the statistical characteristics of continuous and repetitive ECG emissions, indicates that the proposed method has potential and effectiveness to separate weak FECG from multichannel AECG with high precision in high noise condition.
- –
- –
- –
- Machine-learning-based saliency detection and its video decoding application in wireless multimedia communications–
Experimental results show that the proposed method for video-saliency detection is superior to other state-of-the-art saliency-detection methods, either in compressed or uncompressed domain.
- A New 3D Coordinates Acquisition Method on Base of Single Camera–
Experiments show that within1500mm, plane normal direction is better than 0.02° in the measurement accuracy, and can replace stereopair measurement in certain conditions for its higher measurement accuracy.
- The construction of the informationized scientific research administrative platform based on networks–
From the experience of Zhejiang Normal University′s scientific research administrative system exploitation, it was discussed in detail the process of the construction of the informationized scientific research administration based on networks.
- An improved collusion-secure digital ringerprinting scheme for digital video–
This proposed scheme reduces the dependence on mass marking positions, which guarantees the ability of collusion-secure in Dittmann's algorithm, and the capacity of possible copies is increased significantly to the level of practical application.
- A handwritten numeral recognition algorithm based on independent components analysis–
The experiment results show that ICA performs well for feature extraction and this proposed method is more effective in recognizing handwriting character than merely using neural networks directly.
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-15. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.