Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks129 from public data
- 862
- 833
- 621
- Visual Saliency Transformer547
This work develops a novel unified model based on a pure transformer, namely, Visual Saliency Transformer, for both RGB and RGB-D salient object detection (SOD), which takes image patches as inputs and leverages the transformer to propagate global contexts among image patches.
- Salient Object Detection via Integrity Learning480
This work designs a novel integrity learning for SOD, which outperforms the baseline methods in terms of a wide range of metrics and introduces an integrity channel enhancement (ICE) component with the goal of enhancing feature channels that highlight the integral salient objects, while suppressing the other distracting ones.
- CNNs-Based RGB-D Saliency Detection via Cross-View Transfer and Multiview Fusion423
A novel framework based on convolutional neural networks (CNNs), which transfers the structure of the RGB-based deep neural network to be applicable for depth view and fuses the deep representations of both views automatically to obtain the final saliency map is proposed.
- Online Multi-Object Tracking with Dual Matching Attention Networks374
This paper introduces a cost-sensitive tracking loss based on the state-of-the-art visual tracker which encourages the model to focus on hard negative distractors during online learning and proposes Dual Matching Attention Networks (DMAN) with both spatial and temporal attention mechanisms.
- Context-aware Cross-level Fusion Network for Camouflaged Object Detection347
A novel Context-aware Cross-level Fusion Network (C2F-Net) to address the challenging COD task and proposes an Attention-induced Cross- level Fusion Module (ACFM) to integrate the multi-level features with informative attention coefficients.
- 335
- 249
- A Deep Spatial Contextual Long-Term Recurrent Convolutional Network for Saliency Detection207
The proposed DSCLSTM model can significantly boost the saliency detection performance by incorporating both global spatial interconnections and scene context modulation, which may uncover novel inspirations for studies on them in computational saliency models.
- Learning Non-target Knowledge for Few-shot Semantic Segmentation166
A novel framework, namely Non-Target Region Eliminating (NTRE) network, to explicitly mine and eliminate BG and DO regions in the query to improve the model ability of distinguishing the target object from DOs.
- PiCANet: Pixel-Wise Contextual Attention Learning for Accurate Saliency Detection144
This work introduces a novel pixel-wise contextual attention network (PiCANet) to selectively attend to informative context locations at each pixel and introduces the proposed PiCANets into a U-Net model for salient object detection.
- Learning Selective Mutual Attention and Contrast for RGB-D Saliency Detection128
A novel mutual attention model is proposed that uses the non-local attention of one modality to propagate long-range contextual dependencies for the other, thus leveraging complementary attention cues to achieve high-order and trilinear cross-modal interaction.
- 117
- Learning to Predict Eye Fixations via Multiresolution Convolutional Neural Networks110
A novel computation model to predict eye fixations, which adopts a multiresolution convolutional neural network (Mr-CNN) to infer these three types of saliency cues from raw image data simultaneously, demonstrating the superiority of Mr-CNN.
- Weakly Supervised Video Salient Object Detection89
An "Appearance-motion fusion module" and bidirectional ConvLSTM based framework are proposed to achieve effective multi-modal learning and long-term temporal context modeling based on the first weakly supervised video salient object detection model based on relabeled “fixation guided scribble annotations”.
- 86
- 81
- Spatiotemporal Convolutional Neural Network with Convolutional Block Attention Module for Micro-Expression Recognition77
The experimental results show that the 3D CNN with convolutional block attention module outperformed other algorithms in micro-expression recognition.
- Summarize and Search: Learning Consensus-aware Dynamic Convolution for Co-Saliency Detection76
A novel consensus-aware dynamic convolution model is proposed to explicitly and effectively perform the "summarize and search" process to summarize consensus image features and generate dynamic kernels from consensus features to encode the summarized consensus knowledge.
- 75
- FiDeLiS: Faithful Reasoning in Large Language Model for Knowledge Graph Question Answering69
This paper proposes a unified framework, FiDeLiS, designed to improve the factuality of LLM responses by anchoring answers to verifiable reasoning steps retrieved from KGs, and leverages step-wise beam search with a deductive scoring function.
- VST++: Efficient and Stronger Visual Saliency Transformer67
The authors' VST++ model outperforms existing methods while achieving a 25% reduction in computational costs without significant performance compromise, and is evaluated across various transformer-based backbones on RGB, RGB-D, and RGB-T SOD benchmark datasets.
- Latent source mining in FMRI via restricted Boltzmann machine62
The proposed method not only interprets fMRI time courses explicitly to take advantages of deep learning models in latent feature learning but also substantially reduces model complexity and increases the scale of training set to improve training efficiency.
- Light Field Saliency Detection with Dual Local Graph Learning and Reciprocative Guidance56
This paper builds a novel dual graph model to guide the focal stack fusion process using all-focus patterns and introduces a reciprocative guidance scheme to enable mutual guidance between these two kinds of information at multiple steps.
- 55
- 54
- 52
- Harmonizing Performance and Isolation in Microkernels with Efficient Intra-kernel Isolation and Communication51
A redesign of traditional microkernel OSes to harmonize the tension between messaging performance and isolation and a new microkernel called ChCore is designed and implemented based on UnderBridge and also ported to three mainstream microker-nels.
- 49
- CivRealm: A Learning and Reasoning Odyssey in Civilization for Decision-Making Agents46
CivRealm, an environment inspired by the Civilization game, is introduced, providing interfaces for two typical agent types: tensor-based agents that focus on learning, and language-based agents that emphasize reasoning.
- Unsupervised Pre-Training With Language-Vision Prompts for Low-Data Instance Segmentation42
A new method, Unsupervised Pre-training with Language-Vision Prompts (UPLVP), which improves QEIS models’ instance segmentation by bringing language-vision prompts to queries/kernels, and can converge faster and perform better than CNN-based models in low-data regimes.
- 42
- Discriminative Co-Saliency and Background Mining Transformer for Co-Salient Object Detection40
This paper proposes a Discriminative co-salient and background Mining Transformer framework based on several economical multi-grained correlation modules to explicitly mine both co- saliency and background information and effectively model their discrimination.
- Instance-Level Relative Saliency Ranking with Graph Reasoning37
A novel unified model is presented as the first end-to-end solution, where an improved Mask R-CNN is first used to segment salient instances and a saliency ranking branch is then added to infer the relative saliency.
- 36
- 35
- CalibNet: Dual-Branch Cross-Modal Calibration for RGB-D Salient Instance Segmentation33
This study proposes a novel approach for RGB-D salient instance segmentation using a dual-branch cross-modal feature calibration architecture called CalibNet, which simultaneously calibrates depth and RGB features in the kernel and mask branches to generate instance-aware kernels and mask features.
- 30
- Deep learning assisted sound source localization using two orthogonal first-order differential microphone arrays29
Simulation and real-world experimental results show that the proposed deep learning assisted sound localization method can achieve higher spatial resolution and is superior to its state-of-the-art counterparts using the GCC-PHAT or sound intensity features for small-sized arrays in noisy and reverberant environments.
- Towards Instance-adaptive Inference for Federated Learning27
This paper presents a novel FL algorithm, i.e., FedIns, to handle intra-client data heterogeneity by enabling instance-adaptive inference in the FL framework, and resorts to a parameter-efficient fine-tuning method, i.e., scale and shift deep features (SSF) upon a pre-trained model.
- 25
- 25
- 25
- Boosting Low-Data Instance Segmentation by Unsupervised Pre-training with Saliency Prompt24
A new pre-training method that boosts QEIS models by giving Saliency Prompt for queries/kernels and helps QE IS models achieve a similar convergence speed and comparable performance with CNN-based models in low-data regimes is introduced.
- 24
- 23
- RAGNet: Large-Scale Reasoning-Based Affordance Segmentation Benchmark Towards General Grasping22
A large-scale grasping-oriented affordance segmentation benchmark with human-like instructions, named RAGNet, and a comprehensive affordance-based grasping framework, named AffordanceNet, which consists of a VLM pretrained on the authors' massive affordance data and a grasping network that conditions an affordance map to grasp the target.
- 21
- Rethinking feature aggregation for deep RGB-D salient object detection20
This paper proposes another early aggregation scheme to aggregate and propagate multi-modal encoder features at each level, thereby improving the encoder capability and proposes a factorized attention module to efficiently modulate the feature aggregation action for each feature node with multiple learned attention factors.
- 19
- 19
- PA-Net: Learning local features using by pose attention for short-term person re-identification19
A two-stream model named PA-Net is proposed in this paper, which learns the local features through pose attention generated by the pose detection model, which makes the proposed PA- net be more suitable for real-time applications.
- 18
- An End-to-End Pipeline for Early Diagnosis of Acute Promyelocytic Leukemia Based on a Compact CNN Model18
A compact classification model based on attention embedded convolutional neural network blocks is proposed to distinguish promyelocytes from normal leukocytes and outperforms the other evaluated popular classification models owing to its better accuracy and smaller size.
- 17
- A Latent Fingerprint in the Wild Database15
A new wild large-scale latent fingerprint database that includes five different acquisition scenarios: reference fingerprints from optical and capacitive sensors, smartphone fingerprints, latent fingerprints captured from wall surface, Ipad surface, and aluminium foil surface is introduced.
- Multi-grained Temporal Prototype Learning for Few-shot Video Object Segmentation15
This work proposes to leverage multi-grained temporal guidance information for handling the temporal correlation nature of video data, and decomposes the query video information into a clip prototype and a memory prototype for capturing local and long-term internal temporal guidance.
- A deep manifold learning approach for spatial-spectral classification with limited labeled training samples13
The proposed Locality Preserving Convolutional Network relieves the Hughes phenomenon for deep learning, and achieves competitively high classification accuracy compared to other state-of-the-art approaches.
- 12
- Weakly Supervised Ternary Stream Data Augmentation Fine-Grained Classification Network for Identifying Acute Lymphoblastic Leukemia12
A ternary stream-driven weakly supervised data augmentation classification network (WT-DFN) to identify lymphoblasts in a fine-grained scale using microscopic images of peripheral blood smears and yields the best performance on the public dataset and the real clinical dataset among competitive methods.
- 10
- Attention-guided lightweight generative adversarial network for low-light image enhancement in maritime video surveillance10
Experiments indicate that the proposed LGAN can effectively enhance low-light images with better preservation of detail and visual quality than other competing methods.
- 8
- Adapting In-Domain Few-Shot Segmentation to New Domains Without Source Domain Retraining8
The resulting Informative Structure Adaptation (ISA) method effectively addresses domain shifts and equips existing well-trained in-domain FSS models with flexible adaptation capabilities for new domains, eliminating the need to redesign or retrain in-domain FSS models on base data.
- V2VSSC: A 3D Semantic Scene Completion Benchmark for Perception with Vehicle to Vehicle Communication8
This paper proposes the first generalized collaborative SSC framework that allows autonomous vehicles to share sensing information from different sensor views to jointly perform SSC tasks and builds V2VSSC, the first V2V SSC benchmark, on top of the large-scale V2V perception dataset OPV2V.
- MTMU: Multi-domain Transformation based Mamba-UNet designed for unruptured intracranial aneurysm segmentation7
A multi-domain transformation-based Mamba-UNet (MTMU) for UIA segmentation, equipped with the feature encoder consisting of a set of Mamba and Flip blocks, endows the model with the capability of long-range dependency perceiving while balancing computational cost.
- Heterogeneous Value Alignment Evaluation for Large Language Models7
This paper proposes a Heterogeneous Value Alignment Evaluation (HVAE) system, designed to assess the success of aligning LLMs with heterogeneous values, and brings the Social Value Orientation framework from social psychology to examine the behavior of these LLMs.
- 7
- 7
- Evaluation of Customer Behaviour with Machine Learning for Churn Prediction: The Case of Bank Customer Churn in Europe7
The logistic regression classifier is chosen as the baseline model to predict customer churn based on 10,000 customer information of a European bank, and AdaBoost classifier is selected as the prediction model for this case, with recall and AUC score being 0.718 and 0.776.
- AMDFNet: Adaptive multi-level deformable fusion network for RGB-D saliency detection7
An adaptive multi-level deformable fusion network (AMDFNet) is proposed to exploit the cross-modality information to dynamically adjust the boundaries of salient objects by exploring the extra input from another modality.
- Planetary Rover Path Planning Based on Improved A* Algorithm7
An improved A* algorithm is proposed via introducing both the environmental characteristics (such as surface slope and surface toughness) and the rover’s traversability as the constraint conditions and numerical simulations indicate that the improved A*, algorithm has shorter path and higher planning success rate.
- PiCANet: Learning Pixel-wise Contextual Attention in ConvNets and Its Application in Saliency Detection.7
Thorough analyses indicate that the global PiCANet helps to construct global contrast while the local PiCANets help to enhance the feature maps to be more homogenous, thus making saliency detection results more accurate and uniform.
- LLaVA-Endo: a large language-and-vision assistant for gastrointestinal endoscopy6
Experimental results show that LLaVA-Endo demonstrates powerful domain expertise and conversational capabilities, outperforming previous SoTA multimodal methods in the field of GI endoscopy data.
- Uncovering the Redundancy in Graph Self-supervised Learning Models6
It is surprisingly discover that the graph self-supervised learning models are highly redundant at both of neuron and layer levels, which implies that the parameters of graph self-supervised models can be largely reduced, making simultaneously fine-tuning both graph self-supervised learning models and prediction layers more feasible.
- Ontology Model for Situation Awareness of City Tunnel Traffic6
The ontology model for monitoring tunnel traffic situation reflects the basic concepts and semantic meanings of an integrated traffic network and can improve the effectiveness and efficiency of intelligent transportation service.
- 5
- 5
- Deep Hierarchical Interval Type 2 Self-Organizing Fuzzy System for Data-Driven Robot Control5
A new architecture for the fuzzy system, the deep hierarchical self-organizing interval type-2 fuzzy system (DHSOIT2FS), with rules constructed by a rule online update algorithm, consequent parameters updated by iterative least squares, and antecedent parameters are updated using a gradient descent algorithm is proposed.
- TAViS: Text-bridged Audio-Visual Segmentation with Foundation Models4
TAViS is presented, a novel framework that couples the knowledge of multimodal foundation models (ImageBind) for cross-modal alignment and a segmentation foundation model (SAM2) for precise segmentation and achieves superior performance on single-source, multi-source, semantic datasets, and excels in zero-shot settings.
- Semi-supervised medical image segmentation with joint pseudo supervision4
Inspired by the idea of ensemble learning, joint pseudo supervision (JPS) uses two independent teacher models to jointly generate pseudo labels with high-confidence, and then supervises the two student models with it, thereby restricting the accumulation of errors.
- 4
- 4
- 4
- Brain Mapping with High-Resolution fMRI Technology4
Taken together, the fMRI technique and the related informatics tools play an important role in the study of the signal-encoding mechanisms in the olfactory system, which is essential to the understanding of the perception of smell.
- 3
- 3
- 3
- 2
- Unsupervised Learning for Class Distribution Mismatch2
This work proposes Unsupervised Learning for Class Distribution Mismatch (UCDM), which constructs positive-negative pairs from unlabeled data for classifier training and introduces a confidence-based labeling mechanism that iteratively assigns pseudo-labels to valuable real-world data and incorporates them into the training process.
- 2
- Face Recognition from Depth Images with Convolutional Neural Network2
A Convolutional Neural Network structure for face recognition in depth images is proposed and experiments show that the proposed CNN structure can significantly improve the face recognition accuracy, compared with traditional face recognition methods, such as LBP, moment invariant and PCA.
- Artificial Neural Network: Deep or Broad? An Empirical Study2
This paper does a systematic study of depth and breadth of an ANN in terms of its accuracy, bias, variance and convergence performance on 72 standard UCI datasets and argues that broad ANN has better overall performance than deep ANN.
- OdorMapComparer: An Application for Quantitative Analyses and Comparisons of fMRI Brain Odor Maps2
OdorMapComparer provides a tool for quantitative, statistical analyses and comparisons of fMRI odor maps in a fashion that is integrated with the overall odor mapping techniques.
- 2
- 1
- 1
- 1
- A Fast Visual Feature Matching Algorithm in Multi-robot Visual SLAM1
A feature matching algorithm based on map environment is proposed to establish feature libraries by classifying the collected features into two categories during the mobile process of every sub-robot to reduce the feature matching time in visual based multi-ro Bot Simultaneous Localization and Mapping (SLAM).
- An Improved Tsai Approach Using Nonlinear Optimization of Camera Calibration Parameters1
Improved approach to Tsai calibration method is presented, where the image center is updated in the imaging model with radial distortion by minimizing the projection error and the external and internal parameters of the camera are refined by the nonlinear optimization algorithm to optimize the global parameters.
- 1
- 1
- –
- –
- –
- –
- –
- –
- Visual-guided human-object interaction detection–
A VG-HOI model with progressive guidance, query reconstruction, and a conditional uncoupling decoder is proposed to supplement common HOI knowledge and task-specific cues to improve the generalization capability of the model.
- –
- –
- –
- SWU-NIRPV: a near-infrared pose variation face database and pose-invariant face recognition–
An NIR face database with 57 pose variations in a dark environment, which can be used in pose-invariant FR research and showed that a difference of even 10 deg between the gallery and testing sets can dramatically reduce the recognition performance.
- –
- –
- A MapReduce-Based Parallel Mining for Frequent Itemset–
The experimental results show that the MapReduce-based parallel algorithm MR-PrePost+ performs better than PFP in running time, which is one of the best frequent itemset mining parallel algorithms.
- –
- Research on XML Data Structural-Integrity in XML Signature–
A method to protect the structural-integrity of XML digital signature so as to protect XML data element from structural attack through the basic features of traversing binary tree, and ensure the absolute location relationship among elements.
- Research on Intelligent Mobile Platform Base on Monocular Vision and Ultrasonic Sensor–
The intelligent mobile robot can turn smoothly and identify road and obstacles well and navigate stably and achieve the effect of bypassing obstacles through fuzzy controller.
- –
- –
- The Search Tool Based on Competitive Intelligence Analysis–
The author first introduces the tool, and illustrates exactly how to apply this tool into the competitive analysis, and makes a case study for Lenovo.
- –
- –
- Research of Text Segmentation Based on Parallel Genetic Algorithm–
Focusing on the data sparseness of short texts,an algorithm based on knowledge from external corpus is proposed to improve the accuracy of text segmentation, which achieves higher accuracy than the MDA and LDA-based methods in the case of dataSparseness.
- Coding of Peripheral Olfactory Information in the Olfactory Bulb of Small Animals–
To reveal how chemical structure and experimental parameters are represented by the patterns, high resolution functional MRI was used to map the responses of the bulb to various stimuli under different conditions and revealed that different odorants elicited specific and reproducible patterns.
- Studies on the Information Security of IPv6 Internet–
The security mechanism based on IPv6 Internet are briefly introduced at first in the paper, and the main encryption algorithms include MD5,SHA,RSA,ECC, and FAPKC.
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.