Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks136 from public data
- Rethinking Efficient Lane Detection via Curve Modeling223
This paper proposes to exploit the parametric Bézier curve due to its ease of computation, stability, and high freedom degrees of transformations, and proposes the deformable convolution-based feature flip fusion, for exploiting the symmetry properties of lanes in driving scenes.
- DMT: Dynamic mutual training for semi-supervised learning214
This paper proposes mutual training between two different models by a dynamically re-weighted loss function, called Dynamic Mutual Training (DMT), and quantifies inter-model disagreement by comparing predictions from two different models to dynamically re-weight loss in training.
- PromptAD: Learning Prompts with only Normal Samples for Few-Shot Anomaly Detection212
A one-class prompt learning method for few-shot anomaly detection, termed PromptAD, which proposes semantic concatenation which can transpose normal prompts into anomaly prompts by concatenating normal prompts with anomaly suffixes, and introduces the concept of explicit anomaly margin.
- Re-ID Driven Localization Refinement for Person Search160
A differentiable ROI transform layer is developed to effectively transform the bounding boxes from the original images so that the box coordinates can be supervised by the re-ID training other than the original detection task.
- Efficient multimodal large language models: a survey157
This survey summarizes the timeline of representative efficient MLLMs, the current state of research in structures and strategies, and the applications, and the limitations of current efficient MLLM research and promising future directions are discussed.
- 112
- 91
- Unveiling the Power of CLIP in Unsupervised Visible-Infrared Person Re-Identification85
A new prompt learning paradigm for unsupervised visible-infrared person re-identification (USL-VI-ReID) by taking full advantage of the visual-text representation ability from CLIP is proposed and a new memory-swapping contrastive learning is proposed.
- Executive Function Training for Children with Attention Deficit Hyperactivity Disorder83
Although replication with a larger sample and an active control group are needed, EF training program with multiple EF focus and parent involving in real-life activities could be a potentially promising intervention associated with significant EF (near transfer) and ADHD symptoms improvement (far transfer).
- An Adaptive Attention Fusion Mechanism Convolutional Network for Object Detection in Remote Sensing Images82
A convolutional network model with an adaptive attention fusion mechanism (AAFM) based on the backbone network of EfficientDet that improves accuracy and has stronger robustness of the optical image dataset DIOR.
- Semi-Supervised Semantic Segmentation via Dynamic Self-Training and Class-Balanced Curriculum74
The method, Dynamic Self-Training and Class-Balanced Curriculum (DST-CBC), exploits inter-model disagreement by prediction confidence to construct a dynamic loss robust against pseudo label noise, enabling it to extend pseudo labeling to a class-balanced curriculum learning process.
- Boundary-Aware Geometric Encoding for Semantic Segmentation of Point Clouds69
A Boundary Prediction Module (BPM) to predict boundary points and a boundary-aware Geometric Encoding Module (GEM) is designed to encode geometric information and aggregate features with discrimination in a neighborhood, so that the local features belonging to different categories will not be polluted by each other.
- Hierarchical Modeling of Global Context for Document-Level Neural Machine Translation68
This paper proposes a hierarchical model to learn the global context for document-level neural machine translation (NMT), which feedback the extracted global document context to each word in a top-down fashion to distinguish different translations of a word according to its specific surrounding context.
- Frequency-aware Camouflaged Object Detection66
A frequency-aware context aggregation module to suppress high-frequency information and aggregate multi-scale features from a frequency perspective, an adaptive frequency attention module to enhance the features of the learned important frequency components, and a gradient-weighted loss function to guide the proposed method to pay more attention to contour details.
- Omni-supervised Point Cloud Segmentation via Gradual Receptive Field Component Reasoning60
This paper brings the first omni-scale supervision method to point cloud segmentation via the proposed gradual Receptive Field Component Reasoning (RFCR) and brings new state-of-the-art performances for S3DIS as well as Semantic3D and ranks the 1st in the ScanNet benchmark among all the point-based methods.
- 55
- Research and Application of Several Key Techniques in Hyperspectral Image Preprocessing51
This paper focuses on image segmentation, image correction and spatial-spectral dimensional denoising of images in hyperspectral image preprocessing to improve the classification accuracy of hyperspectRAL images.
- Molecular and phenotypic spectrum of Noonan syndrome in Chinese patients51
The genotypic and phenotypic spectra of NS in China are described, providing new insights into distinctive clinical features due to specific pathogenic variant, and ubiquitous abnormal intracranial structure due to SHOC2 pathogenic variants.
- FVNet: 3D Front-View Proposal Generation for Real-Time Object Detection from Point Clouds50
A novel framework called FVNet for 3D front-view proposal generation and object detection from point clouds, which achieves real-time performance with 12ms per point cloud sample and outperforms state-of-the-art techniques which take either camera images or point clouds as input, in terms of accuracy and inference time.
- 46
- One-for-More: Continual Diffusion Model for Anomaly Detection41
An iterative singular value decomposition method based on the transitive property of linear representation, which consumes tiny memory and incurs almost no performance loss, and an anomaly-masked network to enhance the condition mechanism of the diffusion model are proposed.
- GEOcc: Geometrically Enhanced 3D Occupancy Network With Implicit-Explicit Depth Fusion and Contextual Self-Supervision41
GEOcc, a Geometric-Enhanced Occupancy network tailored for vision-only surround-view perception achieves State-of-the-Art performance on the Occ3D-nuScenes dataset with the least image resolution needed and the most weightless image backbone compared with current models, marking an improvement of 3.3% due to the proposed contributions.
- Boosting Night-Time Scene Parsing With Learnable Frequency39
This paper proposes to exploit the image frequency distributions for night-time scene parsing with a Learnable Frequency Encoder (LFE) to model the relationship between different frequency coefficients to measure all frequency components dynamically and proposes a Spatial Frequency Fusion module (SFF) that fuses both spatial and frequency information to guide the extraction of spatial context features.
- 38
- Image Understands Point Cloud: Weakly Supervised 3D Semantic Segmentation via Association Learning32
This paper designs a dual-branch network equipped with an active labeling strategy to maximize the power of tiny parts of labels and to directly realize 2D-to-3D knowledge transfer and establishes a cross-modal self-training framework, which iterates between parameter updating and pseudolabel estimation.
- 31
- Crowd Counting via Multi-layer Regression31
By using multi-layer regression, MRNet achieves significant improvement in crowd counting accuracy, and outperforms the state-of-the-art methods.
- 29
- 29
- A new method proposed to Melanoma-skin cancer lesion detection and segmentation based on hybrid convolutional neural network27
A highly effective Hybrid detection and segmentation method based on the integration of RetinaNet and MaskRCNN, which utilizes a pyramid module of lateral connections and top-down paths to compensate for the loss of spatial features information is proposed.
- Weakly-Supervised Saliency Detection via Salient Object Subitizing27
This paper proposes a model with two modules, Saliency Subitizing Module (SSM) and Saliency Updating Module (SUM), which learns to generate the initial saliency masks using the subitizing information, without the need for any unsupervised methods or some random seeds.
- 24
- Application of intelligent tongue image analysis in Conjunction with microbiomes in the diagnosis of MAFLD21
Combining computer-intelligent tongue diagnosis with microbiome technology enhances MAFLD diagnostic accuracy and provides a convenient early screening reference.
- SceneEncoder: Scene-Aware Semantic Segmentation of Point Clouds with A Learnable Scene Descriptor21
A SceneEncoder module to impose a scene-aware guidance to enhance the effect of global information, which predicts a scene descriptor, which learns to represent the categories of objects existing in the scene and directly guides the point-level semantic segmentation through filtering out categories not belonging to this scene.
- Optimization over Disentangled Encoding: Unsupervised Cross-Domain Point Cloud Completion via Occlusion Factor Manipulation20
This paper disentangle partial scans into three (domain, shape, and occlusion) factors to handle the output gap in cross-domain completion and introduces an optimization stage to maximize the consistency between completed shapes and input scans.
- 20
- Novelty Detection via Contrastive Learning with Negative Data Augmentation20
This work introduces a novel decoder-encoder framework that has significant superiority over cutting-edge novelty detectors and achieves new state-of-the-art results on various novelty detection benchmarks, e.g. CIFAR10 and DCASE.
- Deep multi-center learning for face alignment20
A novel deep learning framework named Multi-Center Learning with multiple shape prediction layers for face alignment that is effective for handling complex occlusions and appearance variations with real-time performance.
- Neuropsychological Profile Related with Executive Function of Chinese Preschoolers with Attention-Deficit/Hyperactivity Disorder19
Preschoolers with ADHD show difficulties of neuropsychological function in many aspects, and the ADHD-combined symptoms (ADHD-C) group had the poorest ratings on inhibition and the ability to Plan/Organize.
- D2U-Net: a dual-path hybrid UNet architecture for precise medical image segmentation17
The proposed D2U-Net is an innovative dual-path Hybrid U-Net architecture specifically implemented to improve feature interactions and optimize the use of multi-scale information, and a Contextual-Spectral Fusion Module (CSFM) specifically developed as a comprehensive information fusion and enhancement approach for encoder–decoder approaches.
- Sketch-to-photo face generation based on semantic consistency preserving and similar connected component refinement15
This work proposes a two-stage sketch-to-photo generative adversarial network for face generation and proposes a semantic loss to maintain semantic consistency and a texture loss to enhance the local fidelity of synthesized images.
- Single Image Deraining via detail-guided Efficient Channel Attention Network15
A novel network single image Deraining via detail-guided Efficient Channel Attention Network (DECAN) to remove rain streaks from rainy images and recover the background of the derained image is proposed.
- MOS: Modeling Object-Scene Associations in Generalized Category Discovery14
The Modeling Object-Scene Associations (MOS) framework is proposed, which utilizes a simple MLP-based scene-awareness module to enhance GCD performance and achieves an exceptional average accuracy improvement of 4% on the challenging fine-grained datasets compared to state-of-the-art methods.
- 13
- 12
- Coupling Context Modeling with Zero Pronoun Recovering for Document-Level Natural Language Generation11
This work proposes a highly adaptive two-stage approach to couple context modeling with ZP recovering to mitigate the ZP problem in NLG tasks and shows that this approach can improve the performance to a great extent.
- MCCH: A novel convex hull prior based solution for saliency detection11
This paper proposes a multi-center convex hull prior based solution for salient object detection that is effective and competitive, compared against state-of-the-art saliency detection algorithms.
- Time-Domain Analysis of Noise Coupling Between Package and PCB Power/Ground Planes Based on WLP-FDTD11
- A parallelized screen-based method for rendering polylines and polygons on terrain surfaces11
This research proposes a new screen-space based method to render solid color 2D polylines and polygons on terrain that eliminates undesirable rendering artifacts and a parallel computing schema was proposed and implemented to improve rendering performance.
- 10
- LLaVA-VSD: Large Language-and-Vision Assistant for Visual Spatial Description10
A Large Language-and-Vision Assistant for Visual Spatial Description, named LLaVA-VSD, which is designed for the classification, description, and open-ended description of visual spatial relationships and demonstrates excellent multimodal conversational capabilities.
- Textual Decomposition Then Sub-motion-space Scattering for Open-Vocabulary Motion Generation10
This paper proposes to leverage the atomic motion as an intermediate representation, and leverage two orderly coupled steps, i.e., Textual Decomposition and Sub-motion-space Scattering, to address the full mapping problem of the open-vocabulary motion generation.
- 10
- 10
- PIG: Prompt Images Guidance for Night-Time Scene Parsing9
This paper proposes a Night-Focused Network (NFNet) to learn night-specific features from both target domain images and prompt images and proposes two data augmentation strategies: the Prompt Mixture Strategy (PMS) and the Alternate Mask Strategy (AMS), aimed at mitigating the overfitting of the NFNet to a few prompt images.
- 9
- 9
- A Two-Step Simulated Annealing Algorithm for Spectral Data Feature Extraction8
The experimental results show that compared with the traditional feature extraction method, the accuracy indexes of the inversion model established by using the TSSA algorithm for feature extraction were significantly improved, with the determination coefficient R2 of 0.9654.
- 7
- 7
- Image-text Retrieval with Main Semantics Consistency7
This paper proposes a semantically optimized approach with a novel Main Semantics Consistency (MSC) loss function, which aims to rank the semantically most similar images (or texts) corresponding to the given query at the top position during the retrieval process.
- 7
- LW‐CovidNet: Automatic covid‐19 lung infection detection from chest X‐ray images7
A novel covid‐19 lightweight CNN(LW‐CovidNet) method that surpasses most cutting‐edge detection methods and also contributes to the advancement of state‐of‐the‐art performance is proposed.
- 6
- 6
- 6
- Towards Discourse-Aware Document-Level Neural Machine Translation6
This work proposes a document-level NMT system enhanced with the discourse-aware document context, which is named Disco2NMT, which significantly surpasses both context-agnostic and context-aware baseline systems on multiple evaluation indicators.
- 6
- 5
- Document-Level Neural Machine Translation with Hierarchical Modeling of Global Context5
A hierarchical model to learn the global context for document-level neural machine translation (NMT) is proposed through a sentence encoder to capture intra-sentence dependencies and a document encoders to model document- level inter-sentences consistency and coherence.
- 5
- Detecting and Translating Dropped Pronouns in Neural Machine Translation5
This work proposes a method to both automatically detect the dropped pronouns (DPs) and recover their translation equivalences rather than their original forms in source sentences, simultaneously performed as a sequence labeling task on source sentences.
- 4
- Domain-Hallucinated Updating for Multi-Domain Face Anti-spoofing4
This work proposes a novel Domain-Hallucinated Updating (DHU) framework to facilitate the hallucination of data for updating FAS model and demonstrates that the proposed method outperforms state-of-the-art competitors in terms of effectiveness.
- 4
- Night-time Semantic Segmentation with a Large Real Dataset.4
Extensive experiments show that training on NightCity can significantly improve the performance of night-time semantic segmentation and that the exposure-aware model outperforms the state-of-the-art segmentation methods, yielding top performances on the authors' benchmark dataset.
- Rendering 2D Lines on 3D Terrain Model with Optimization in Visual Quality and Running Performance4
This work proposes a novel screen‐based method for rendering 2D vector lines with the accuracy of more than one pixel on the screen in real time, realized in a heterogeneously parallel model so that the performance is improved and becomes acceptable.
- 3
- 3
- PFDepth: Heterogeneous Pinhole-Fisheye Joint Depth Estimation via Distortion-aware Gaussian-Splatted Volumetric Fusion3
This is the first systematic study of heterogeneous pinhole-fisheye depth estimation, offering both technical novelty and valuable empirical insights, and sets a state-of-the-art performance on KITTI-360 and RealHet datasets over current mainstream depth networks.
- Learnable scene prior for point cloud semantic segmentation3
A SceneEncoder module to give scene-aware guidance to final segmentation results, which learns to predict a scene descriptor that represents the categories existing in the scene and uses it to filter out categories not belonging to this scene directly.
- 3
- 3D Visualization of Trees Based on a Sphere-Board Model3
The results show that the sphere-board-based tree model (SBTM) can achieve a better compromise between realism and performance than can the BBTM or CPTM and can better support lighted effects because of its cubic geometrical features.
- 3
- Self-Aware Safety Augmentation: Leveraging Internal Semantic Understanding to Enhance Safety in Vision-Language Models2
Self-Aware Safety Augmentation (SASA) is proposed, a technique that projects informative semantic representations from intermediate layers onto earlier safety-oriented layers and leverages the model's inherent semantic understanding to enhance safety recognition without fine-tuning.
- 2
- 2
- 2
- 2
- 2
- HSNet: hierarchical semantics network for scene parsing2
This paper aims to introduce the HSP into scene parsing, by proposing a hierarchical semantics network (HSNet), which enables us to learn multi-level, hierarchy-aware features via forward feature transfer and backward feature regularization.
- A Recursive Information Flow Gated Model for RST-Style Text-Level Discourse Parsing2
This paper proposes a recursive model to fuse multiple information flows and strengthen the representation of text spans through shift-reduce operations, and shows that the proposed parser can outperform the state-of-the-art in nuclearity detection under stringent discourse parsing evaluations.
- RST Discourse Parsing with Tree-Structured Neural Networks2
A tree-structured neural network is built for RST discourse parsing and two tracking LSTMs are introduced to store long-distance information of a document to strengthen the representations for sentences and the entire document.
- 2
- [Effects of sleep deprivation on the intelligence structure of school-age children in Changsha, China].2
Sleep deprivation adversely affects intelligence development, especially VIQ, in primary school students, and the adverse effects of sleep deprivation are mainly seen in students with moderate and severe sleep deprivation.
- 1
- Glass Surface Detection Method Based on Visual Distortion1
The proposed method achieves improvements in IoU ranging from 0.83% to 4.73%, improvements in F -measure ranging from 1.40% to 6.60%, reductions in MAE ranging from 0.72% to 2.60%, and decreases in BER ranging from 0.58% to 2.66% across four benchmark datasets.
- 1
- 1
- 1
- 1
- 1
- 1
- Domain Alignment with Large Vision-language Models for Cross-domain Remote Sensing Image Retrieval1
A dual-modality prototype guided pseudo-labeling mechanism, which leverages the pre-trained large vision-language model to assign pseudo-labels for all unlabeled source domain images and target domain images and a loss reweighting strategy which incorporates the confidence scores as weight values into the contrastive loss to mitigate the impact of noisy pseudo-labels.
- 1
- 1
- Top-down Text-Level Discourse Rhetorical Structure Parsing with Bidirectional Representation Learning1
It is argued that combining local and global information for discourse parsing is more sensible, and a top-down discourse parser with bidirectional representation learning capabilities is introduced, to prove this.
- 1
- EDTC: A Corpus for Discourse-Level Topic Chain Parsing1
This research presents insight on discourse-level topic chain (DTC) parsing which aims at discovering new topics and in-vestigating how these topics evolve over time within an article to address the lack of data.
- 1
- A Shape-Aware Feature Extraction Module for Semantic Segmentation of 3D Point Clouds1
A Shape-Aware Feature Extraction (SAFE) module is proposed that explicitly describes the spatial distribution of points in the neighborhood by well-designed distribution descriptors and replaces the conventional solid neighborhood with a hollow spherical neighborhood.
- Human-Centric Visual Relation Segmentation Using Mask R-CNN and VTransE1
A novel human-centric visual relation segmentation method based on Mask R-CNN model and VTransE model, which outperforms other methods in Person in Context Challenge.
- 1
- 1
- –
- –
- –
- Learning Pulse Image with Deep Dynamic Frequency Network for Cardiovascular Diseases Diagnosis–
This paper improves the original equipment and utilizes array sensors to establish two array pulse signal datasets and benchmarks for identifying CVD, including hypertension and coronary heart disease, and proposes a deep dynamic frequency network (DDFNet) to model the frequency distribution of pulse signals for distinguishing CVD.
- Neural networks recruited for numerical rule learning and application–
Results demonstrate how multiple neural networks underlying both numeric and nonnumeric processing can interact to allow people to make mathematical decisions.
- –
- –
- –
- –
- Isolation and Integration: A Strong Pre-trained Model-Based Paradigm for Class-Incremental Learning–
This work proposes a brand-new stage-isolation based class incremental learning framework, which leverages parameter-efficient tuning technique to finetune the pre-trained model for each task, thus mitigating information interference and conflicts among tasks.
- –
- Design and Research of Intelligent QA System for Flight Crew Operating Manual–
Based on the comparison of response speed and retrieval capability, the best model combination is obtained for the flight crew operation manual dataset, and suggestions are made for the model-related performance improvement.
- Coreference Resolution: Are the Eliminated Spans Totally Worthless?–
Experimental results on the document-level CoNLL-2012 Shared Task English dataset show that eliminated spans are indeed much effective and the proposed mention representation refining strategy can achieve competitive results when compared with previous state-of-the-art in coreference resolution.
- –
- –
- –
- Object-Level Salience Detection by Progressively Enhanced Network–
This framework is composed of two main components: progressively enhanced network (PEN) for amplifying the specified layers of the network and merging the global context simultaneously; (2) object-level boundary extraction module (OBEM) for extracting the complete boundary of the salient object.
- –
- Facial Landmark Detection Under Large Pose–
A two-stage cascade regression framework using patch-difference features to overcome the above problem of facial landmark detection and achieves superior results on COFW dataset.
- –
- An Intelligent Surveillance System for Crowed Abnormal Detection–
An intelligent surveillance system for emergency detection is designed for campus and an edge detection is employed to eliminate the effect of brightness and divide the image into limited layers to detect separately after edge detection, based on pixel features.
- –
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.