Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks338 from public data
- 3,850
- Res2Net: A New Multi-Scale Backbone Architecture3,450
This paper proposes a novel building block for CNNs, namely Res2Net, by constructing hierarchical residual-like connections within one single residual block that represents multi-scale features at a granular level and increases the range of receptive fields for each network layer.
- Attention mechanisms in computer vision: A survey2,538
This survey provides a comprehensive review of various attention mechanisms in computer vision and categorize them according to approach, such as channel attention, spatial attention, temporal attention, and branch attention.
- 2,466
- Structure-Measure: A New Way to Evaluate Foreground Maps1,970
A novel, efficient, and easy to calculate measure known as S-measure (structural measure) to evaluate foreground maps, which simultaneously evaluates region-aware and object-aware structural similarity between a foreground map and a ground-truth map.
- 1,938
- Enhanced-alignment Measure for Binary Foreground Map Evaluation1,716
This paper takes a detailed look at current binary FM evaluation measures and proposes a novel and effective E-measure (Enhanced-alignment measure), jointly capturing image-level statistics and local pixel matching information, which demonstrates the superiority of this measure over the available measures on 4 popular datasets via 5 meta-measures.
- Deeply Supervised Salient Object Detection with Short Connections1,458
This paper proposes a new salient object detection method by introducing short connections to the skip-layer structures within the HED architecture, which takes full advantage of multi-level and multi-scale features extracted from FCNs, providing more advanced representations at each layer, a property that is critically needed to perform segment detection.
- Richer Convolutional Features for Edge Detection1,240
RCF encapsulates all convolutional features into more discriminative representation, which makes good usage of rich feature hierarchies, and is amenable to training via backpropagation, and achieves state-of-the-art performance on several available datasets.
- EGNet: Edge Guidance Network for Salient Object Detection1,212
This paper presents an edge guidance network (EGNet) for salient object detection with three steps to simultaneously model these two kinds of complementary information in a single network to solve the complementarity between salient edge information and salient object information.
- A Simple Pooling-Based Design for Real-Time Salient Object Detection1,145
This work solves the problem of salient object detection by investigating how to expand the role of pooling in convolutional neural networks by building a global guidance module (GGM) and designing a feature aggregation module (FAM) to make the coarse-level semantic information well fused with the fine-level features from the top-down path- way.
- Visual attention network1,121
This paper proposes a novel linear attention named large kernel attention (LKA) to enable self-adaptive and long-range correlations in self-attention while avoiding its shortcomings, and presents a neural network based on LKA, namely Visual Attention Network (VAN).
- 1,085
- 1,021
- 939
- 915
- Object Region Mining with Adversarial Erasing: A Simple Classification to Semantic Segmentation Approach900
This work investigates a principle way to progressively mine discriminative object regions using classification networks to address the weakly-supervised semantic segmentation problems and proposes a new adversarial erasing approach for localizing and expanding object regions progressively.
- Large Selective Kernel Network for Remote Sensing Object Detection850
This paper considers these priors and proposes the lightweight Large Selective Kernel Network (LSKNet), which can dynamically adjust its large spatial receptive field to better model the ranging context of various objects in remote sensing scenarios.
- Strip Pooling: Rethinking Spatial Pooling for Scene Parsing817
Beyond conventional spatial pooling that usually has a regular shape of NxN, a new pooling strategy is introduced, called strip pooling, which considers a long but narrow kernel, i.e., 1xN or Nx1.
- 790
- Concealed Object Detection786
The first systematic study on concealed object detection (COD), which aims to identify objects that are visually embedded in their background, is presented, and rich annotations including object categories, object boundaries, challenging attributes, object-level labels, and instance-level annotations are provided.
- Rethinking RGB-D Salient Object Detection: Models, Data Sets, and Large-Scale Benchmarks729
It is demonstrated that D3Net can be used to efficiently extract salient object masks from real scenes, enabling effective background-changing application with a speed of 65 frames/s on a single GPU.
- Sketch2Photo706
A system that composes a realistic picture from a simple freehand sketch annotated with text labels, able to automatically select suitable photographs to generate a high quality composition, using a filtering scheme to exclude undesirable images.
- Low-Light Image and Video Enhancement Using Deep Learning: A Survey651
This paper proposes a low-light image and video dataset, in which the images and videos are taken by different mobile phones’ cameras under diverse illumination conditions, and provides a unified online platform that covers many popular LLIE methods.
- 638
- STC: A Simple to Complex Framework for Weakly-Supervised Semantic Segmentation617
A simple to complex (STC) framework in which only image-level annotations are utilized to learn DCNNs for semantic segmentation, which demonstrates the superiority of the proposed STC framework compared with other state-of-the-arts frameworks.
- The Visual Object Tracking VOT2014 Challenge Results606
The Visual Object Tracking challenge 2014, VOT2014, aims at comparing short-term single-object visual trackers that do not apply pre-learned models of object appearance and introduces a new dataset with full annotation of targets by rotated bounding boxes and per-frame attribute.
- 597
- 542
- 537
- 503
- 442
- Review of Visual Saliency Detection With Comprehensive Information434
Different types of saliency detection algorithms are reviewed, the important issues of the existing methods are summarized, the existent problems and future works are discussed, and the experimental analysis and discussion are conducted to provide a holistic overview of differentSaliency detection methods.
- PhotoMaker: Customizing Realistic Human Photos via Stacked ID Embedding410
PhotoMaker is introduced, an efficient personalized text-to-image generation method, which mainly encodes an arbitrary number of input ID images into a stack ID embed-ding for preserving ID information, and an ID-oriented data construction pipeline to assemble the training data.
- 373
- Dense Attention Fluid Network for Salient Object Detection in Optical Remote Sensing Images355
An end-to-end Dense Attention Fluid Network (DAFNet) for SOD in optical RSIs that significantly outperforms the existing state-of-the-art SOD competitors is proposed and a new and challenging optical RSI dataset is constructed that contains 2,000 images with pixel-wise saliency annotations.
- MS-TCN++: Multi-Stage Temporal Convolutional Network for Action Segmentation352
A multi-stage architecture for the temporal action segmentation task that overcomes the limitations of the previous approaches and achieves state-of-the-art results on three datasets: 50Salads, Georgia Tech Egocentric Activities (GTEA), and the Breakfast dataset.
- Salient Objects in Clutter: Bringing Salient Object Detection to the Foreground349
This work identifies a serious design bias of existing SOD datasets which assumes that each image contains at least one clearly outstanding salient object in low clutter, and proposes a new high quality dataset and updates the previous saliency benchmark.
- 336
- Masked Diffusion Transformer is a Strong Image Synthesizer329
A Masked Diffusion Transformer (MDT) is proposed that introduces a mask latent modeling scheme to explicitly enhance the DPMs’ ability to contextual relation learning among object semantic parts in an image.
- RGB-D salient object detection: A survey329
This paper provides a comprehensive survey of RGB-D based salient object detection models from various perspectives, and review related benchmark datasets in detail, and investigates the ability of existing models to detect salient objects.
- P2T: Pyramid Pooling Transformer for Scene Understanding321
This paper proposes to adapt pyramid pooling to Multi-Head Self-Attention (MHSA) in the vision transformer, simultaneously reducing the sequence length and capturing powerful contextual features, in a universal vision transformer backbone, dubbed Pyramid Pooling Transformer (P2T).
- Revisiting Video Saliency: A Large-Scale Benchmark and a New Model311
This work introduces a new benchmark for predicting human eye movements during dynamic scene free-viewing, and proposes a novel video saliency model that augments the CNN-LSTM network architecture with an attention mechanism to enable fast, end-to-end saliency learning.
- 311
- SalientShape: group saliency in image collections309
This work introduces group saliency to achieve superior unsupervised salient object segmentation by extracting salient objects (in collections of pre-filtered images) that maximize between-image similarities and within-image distinctness.
- Conv2Former: A Simple Transformer-Style ConvNet for Visual Recognition290
A deep look at the internal structure of self-attention is taken and a simple Transformer style convolutional neural network (ConvNet) is presented for visual recognition and a family of hierarchical ConvNets using the proposed convolutional modulation, termed Conv2Former is built.
- Image Inpainting With Learnable Bidirectional Attention Maps287
This paper presents a learnable attention map module for learning feature re-normalization and mask-updating in an end-to-end manner, which is effective in adapting to irregular holes and propagation of convolution layers.
- EDN: Salient Object Detection via Extremely-Downsampled Network284
This work introduces an Extremely-Downsampled Network (EDN), which employs an extreme downsampling technique to effectively learn a global view of the whole image, leading to accurate salient object localization and construct the Scale-Correlated Pyramid Convolution (SCPC) to accomplish better multi-level feature fusion.
- 275
- 273
- 265
- 260
- Deep Hough Transform for Semantic Line Detection248
This paper incorporates the classical Hough transform technique into deeply learned representations and proposes a one-shot end-to-end learning framework for line detection by parameterizing lines with slopes and biases, which transforms the problem of detecting semantic lines in the spatial domain into spotting individual points in the parametric domain.
- YOLO-MS: Rethinking Multi-Scale Representation Learning for Real-Time Object Detection246
The core design is based on a series of investigations on how multi-branch features of the basic block and convolutions with different kernel sizes affect the detection performance of objects at different scales, and the outcome is a new strategy that can significantly enhance multi-scale feature representations of real-time object detectors.
- A Shape‐Preserving Approach to Image Resizing245
A novel image resizing method which attempts to ensure that important local regions undergo a geometric similarity transformation, and at the same time, to preserve image edge structure, and is efficient, and offers a closed form solution.
- LSKNet: A Foundation Lightweight Backbone for Remote Sensing237
This paper considers the significance of the identified priors and proposes a lightweight Large Selective Kernel Network (LSKNet) backbone, which can dynamically adjust its large spatial receptive field to better model the ranging context of various objects in remote sensing scenarios.
- Towards An End-to-End Framework for Flow-Guided Video Inpainting234
An End-to-End framework for Flow-Guided Video Inpainting (E2 FGVI) through elaborately designed three trainable modules, namely, flow completion, feature propagation, and content hallucination modules that can be Jointly optimized, leading to a more efficient and effective inpainting process.
- AMT: All-Pairs Multi-Field Transforms for Efficient Frame Interpolation229
This work presents All-Pairs Multi-Field Transforms (AMT), a new network architecture for video frame interpolation based on two essential designs that derive multiple groups of fine-grained flow fields from one pair of updated coarse flows for performing backward warping on the input frames separately.
- 227
- Unsupervised Scale-Consistent Depth Learning from Video226
A monocular depth estimation method SC-Depth, which requires only unlabelled videos for training and enables the scale-consistent prediction at inference time and the proposed hybrid Pseudo-RGBD SLAM shows compelling results in KITTI, and it generalizes well to the KAIST dataset without additional training.
- Highly Efficient Salient Object Detection with 100K Parameters222
This paper proposes a flexible convolutional module, namely generalized OctConv (gOctConv), to efficiently utilize both in-stage and cross-stages multi-scale features, while reducing the representation redundancy by a novel dynamic weight decay scheme.
- RepFinder213
A novel framework where user scribbles are used to guide detection and extraction of repeated elements, which robustly extracts the repetitions along with their deformations and demonstrates the versatility of the framework on a large set of inputs of varying complexity.
- Localization Distillation for Dense Object Detection208
This paper reformulates the knowledge distillation process on localization to present a novel localization distillation method which can efficiently transfer the localization knowledge from the teacher to the student and heuristically introduces the concept of valuable localization region that can aid to selectively distill the semantic and localization knowledge for a certain region.
- 202
- CamoFormer: Masked Separable Attention for Camouflaged Object Detection195
Inspired by the multi-head self-attention in Transformers, a simple masked separable attention (MSA) for camouflaged object detection is presented and proposed to capture high-resolution semantic representations progressively based on a simple top-down decoder with the proposed MSA to attain precise segmentation results.
- 193
- Pyramid Constrained Self-Attention Network for Fast Video Salient Object Detection184
A Constrained Self-Attention (CSA) operation to capture motion cues, based on the prior that objects always move in a continuous trajectory is designed, which outperforms previous state-of-the-art methods in both accuracy and speed.
- 181
- 177
- MobileSal: Extremely Efficient RGB-D Salient Object Detection176
This article proposes an implicit depth restoration (IDR) technique to strengthen the mobile networks’ feature representation capability for RGB-D SOD, and proposes compact pyramid refinement (CPR) for efficient multi-level feature aggregation to derive salient objects with clear boundaries.
- 171
- Re-thinking Co-Salient Object Detection170
A novel CoEG-Net is proposed that augments the authors' prior model EGNet with a co-attention projection strategy to enable fast common information learning and fully leverages previous large-scale SOD datasets and significantly improves the model scalability and stability.
- 165
- Dynamic Feature Integration for Simultaneous Detection of Salient Object, Edge, and Skeleton162
Though these tasks are naturally quite different, the proposed network can work well on all of them and even perform better than current single-purpose state-of-the-art methods.
- CrossKD: Cross-Head Knowledge Distillation for Object Detection160
This paper presents a general and effective prediction mimicking distillation scheme, called CrossKD, which delivers the intermediate features of the student's detection head to the teacher's detection head, and the resulting crosshead predictions are forced to mimic the teacher's predictions.
- Representation Compensation Networks for Continual Semantic Segmentation155
This work proposes to use a structural re-parameterization mechanism, named representation compensation (RC) module, to decouple the representation learning of both old and new knowledge, and outperforms state-of-the-art performance.
- Large-Scale Unsupervised Semantic Segmentation153
This work proposes a new problem of large-scale unsupervised semantic segmentation (LUSS) with a newly created benchmark dataset to help the research progress and presents a simple yet effective method that works surprisingly well for LUSS.
- 151
- Advances in deep concealed scene understanding148
A comprehensive survey of deep learning techniques aimed at CSU, including a taxonomy, task-specific challenges, and ongoing developments is presented, and the largest and latest benchmark for concealed object segmentation (COS) is offered.
- 148
- DFormer: Rethinking RGBD Representation Learning for Semantic Segmentation144
Experimental results show that the DFormer achieves new state-of-the-art performance on these two tasks with less than half of the computational cost of the current best methods on two RGB-D semantic segmentation datasets and five RGB-D salient object detection datasets.
- 142
- 137
- 135
- Gradient-Induced Co-Saliency Detection135
A gradient-induced co-saliency detection (GICD) method that first abstracts a consensus representation for the grouped images in the embedding space; then, by comparing the single image with consensus representation, utilizes the feedback gradient information to induce more attention to the discriminative co- salient features.
- SAM struggles in concealed scenes — empirical study on “Segment Anything”134
This report chooses three concealed scenes, i.e., camouflaged animals, industrial defects, and medical lesions, to evaluate SAM under unprompted settings, and observation is that SAM looks unskilled in concealed scenes.
- 132
- 132
- 131
- Temporal Modulation Network for Controllable Space-Time Video Super-Resolution130
A Temporal Modulation Network (TMNet) to interpolate arbitrary intermediate frame(s) with accurate high-resolution reconstruction and demonstrates that the TMNet outperforms previous STVSR methods.
- 123
- 117
- 117
- Semantic Edge Detection with Diverse Deep Supervision113
A novel fully convolutional neural network using diverse deep supervision within a multi-task framework where bottom layers aim at generating category-agnostic edges, while top layers are responsible for the detection of category-aware semantic edges is proposed.
- DNA: Deeply Supervised Nonlinear Aggregation for Salient Object Detection113
The proposed saliency detector, a modified U-Net architecture with DNA, performs favorably against state-of-the-art methods on various datasets and evaluation metrics without bells and whistles and can successfully break through the bottleneck of the current linear approaches.
- Associating Inter-image Salient Instances for Weakly Supervised Semantic Segmentation113
This paper uses an instance-level salient object detector to automatically generate salient instances (candidate objects) for training images, and proposes a graph-partitioning-based clustering algorithm that outperforms state-of-the-art weakly supervised alternatives by a large margin.
- Interactive images113
This work's interactive framework creates partial scene reconstructions based on cuboid-proxies with minimal user interaction and allows a range of intuitive image edits mimicking real-world behavior, which are otherwise difficult to achieve.
- Improved Protein Structure Prediction Using a New Multi‐Scale Network and Homologous Templates110
trRosettaX, an improved version of trRosetta for protein structure prediction is presented, which suggests the effectiveness of using the multi‐scale network and the benefit of incorporating homologous templates into the network.
- 110
- 110
- CorrMatch: Label Propagation via Correlation Matching for Semi-Supervised Semantic Segmentation109
This paper proposes to conduct pixel propagation by modeling the pairwise similarities of pixels to spread the high-confidence pixels and dig out more, and performs region propagation to enhance the pseudo labels with accurate class-agnostic masks extracted from the correlation maps.
- 109
- SemanticPaint: Interactive 3D Labeling and Learning at your Fingertips104
SemanticPaint allows users to simultaneously scan their environment whilst interactively segmenting the scene simply by reaching out and touching any desired object or surface, and continuously learns from these segmentations, and labels new unseen parts of the environment.
- SemanticPaint104
SemanticPaint allows users to simultaneously scan their environment whilst interactively segmenting the scene simply by reaching out and touching any desired object or surface, and continuously learns from these segmentations, and labels new unseen parts of the environment.
- 100
- Localization Distillation for Object Detection99
This paper presents a novel localization distillation (LD) method which can efficiently transfer the localization knowledge from the teacher to the student and introduces the concept of valuable localization region that can aid to selectively distill the classification and localization knowledge for a certain region.
- Class-Specific Semantic Reconstruction for Open Set Recognition98
A novel method, called Class-Specific Semantic Reconstruction (CSSR), that integrates the power of AE and prototype learning, which achieves outstanding performance in both close and open set recognition and is sufficiently simple and flexible to incorporate into existing frameworks.
- 98
- Faster Diffusion: Rethinking the Role of UNet Encoder in Diffusion Models97
It is found that encoder features change gently, whereas the decoder features exhibit substantial variations across different time-steps, which inspired us to conduct the first comprehensive study of the UNet encoder.
- S4Net: Single stage salient-instance segmentation97
This paper designs a single stage salient instance segmentation framework, with a novel segmentation branch that regards not only local context inside each detection window but also the surrounding context, enabling us to distinguish instances in the same scope even with partial occlusion.
- Conditional Variational Image Deraining96
A Conditional Variational Image Deraining (CVID) network is proposed for better deraining performance, leveraging the exclusive generative ability of Conditionalvariational Auto-Encoder on providing diverse predictions for the rainy image.
- 96
- Strip R-CNN: Large Strip Convolution for Remote Sensing Object Detection95
Strip R-CNN is the first work to systematically explore large strip convolutions for remote sensing object detection, finding that strip convolutions enable directional feature aggregation along the dominant spatial dimension of slender objects, reducing background interference while preserving essential geometric information.
- Global2Local: Efficient Structure Search for Video Action Segmentation92
A global-to-local search scheme that exploits both global search to find the coarse combinations and local search to get the refined receptive field combination patterns further and an expectation guided iterative local search scheme to refine combinations effectively.
- Intelligent Visual Media Processing: When Graphics Meets Vision92
Recent research on how computer vision techniques benefit computer graphics techniques and vice versa is surveyed, and research on analysis, manipulation, synthesis, and interaction is covered.
- DenseCut: Densely Connected CRFs for Realtime GrabCut91
A very fast segmentation technique which still achieves very high quality results is demonstrated and is proposed to replace the time consuming iterative refinement of global colour models in traditional GrabCut formulation by a densely connected crf.
- Internet visual media processing: a survey with graphics and vision applications89
Recent research as regards processing of large collections of images and video, including work on analysis, manipulation, and synthesis is surveyed, and possible future directions in this emerging research area are suggested.
- 87
- Bilateral Functions for Global Motion Modeling87
This paper uses the bilateral domain to reformulate a piecewise smooth constraint as continuous global modeling constraint and demonstrates how the model can reliably obtain large numbers of good quality correspondences over wide baselines, while keeping outliers to a minimum.
- Optimizing the F-Measure for Threshold-Free Salient Object Detection86
The relaxed F-measure is proposed which is differentiable w.r.t the posterior and can be easily appended to the back of CNNs as the loss function, named FLoss, which can continuously force the network to produce polarized activations.
- ICNet: Intra-saliency Correlation Network for Co-Saliency Detection84
An Intra-saliency Correlation Network (ICNet) is proposed to extract intra-saliency cues from the single image saliency maps (SISMs) predicted by any off-the-shelf SOD method, and obtain inter-saliency cues by correlation techniques.
- Depth-aware neural style transfer82
A novel approach for neural style transfer is introduced that integrates depth preservation as additional loss, preserving overall image layout while performing style transfer, and producing an image that preserves its semantic content.
- 81
- 79
- DFormerv2: Geometry Self-Attention for RGBD Semantic Segmentation77
This paper investigates a new way to learn RGBD feature representations and presents DFormerv2, a strong RGBD encoder that explicitly uses depth maps as geometry priors rather than encoding depth information with neural networks.
- ImageSpirit72
This article proposes treating nouns as object labels and adjectives as visual attribute labels to formulate the image parsing problem as one of jointly estimating per-pixel object and attribute labels from a set of training images, and proposes an efficient (interactive time) solution.
- 71
- HFS: Hierarchical Feature Selection for Efficient Image Segmentation69
This paper proposes a real-time system, Hierarchical Feature Selection (HFS), that performs image segmentation at a speed of 50 frames-per-second, and can be used in a variety computer vision tasks that are built on top ofimage segmentation and superpixel extraction.
- 68
- 68
- 66
- Salient Objects in Clutter65
It is argued that improving the dataset can yield higher performance gains than focusing only on the decoder design, and several dataset-enhancement strategies are investigated, including label smoothing to implicitly emphasize salient boundaries, random image augmentation to adapt saliency models to various scenarios, and self-supervised learning as a regularization strategy to learn from small datasets.
- 64
- 62
- DOTS: Decoupling Operation and Topology in Differentiable Architecture Search60
The proposed Decouple the Operation and Topology Search (DOTS), which decouples the topology representation from operation weights and makes an explicit topology search, and is an effective solution for differentiable NAS.
- Co-Salient Object Detection with Co-Representation Purification57
This paper proposes a Co-Representation Purification Purification (CoRP) method aiming at searching noise-free co-representation, and uses the prediction to iteratively reduce irrelevant embeddings in the co- representation.
- RF-Next: Efficient Receptive Field Search for Convolutional Neural Networks57
This work proposes to find better receptive field combinations through a global-to-local search scheme, and proposes an expectation-guided iterative local search scheme to refine combinations effectively.
- FLIC: Fast linear iterative clustering with active search56
This paper proposes a novel search algorithm called “active search” which explicitly considers neighbor continuity, and designs a back-and-forth traversal strategy and a joint assignment and update step to speed up the algorithm.
- 55
- Hi-Fi: Hierarchical Feature Integration for Skeleton Detection55
A new convolutional neural network architecture is presented by introducing a novel hierarchical feature integration mechanism, named Hi-Fi, to address the object skeleton detection problem and significantly outperforms the state-of-the-art methods in terms of effectively fusing features from very different scales.
- One-Prompt-One-Story: Free-Lunch Consistent Text-to-Image Generation Using a Single Prompt54
This paper proposes a novel training-free method for consistent text-to-image (T2I) generation, termed "One-Prompt-One-Story", which concatenates all prompts into a single input for T2I diffusion models, initially preserving character identities.
- Weakly Supervised Semantic Segmentation via Alternate Self-Dual Teaching54
A novel end-to-end learning framework, alternate self-dual teaching (ASDT), based on a dual-teacher single-student network architecture is built, inspired by the Pulse Width (PW) modulation, to alleviate the influence of the imperfect knowledge from either teacher model on the KD process.
- DEL: Deep Embedding Learning for Efficient Image Segmentation53
A novel method called DEL (deep embedding learning) which can efficiently transform superpixels into image segmentation which can achieve comparable segments when compared with MCG but is much faster than it, i.e. 11.4fps vs. 0.07fps.
- Faster Diffusion: Rethinking the Role of the Encoder for Diffusion Model Inference52
This paper conducts a comprehensive study of the UNet encoder and empirically analyze the encoder features, which provides insights regarding their changes during the inference process, and finds that encoder features change minimally, whereas the decoder features exhibit substantial variations across different time-steps.
- Ordered or Orderless: A Revisit for Video Based Person Re-Identification52
This paper presents a simple yet surprisingly powerful approach for VPRe-id, where it is shown that the common practice of employing recurrent neural networks to aggregate temporal-spatial features may not be optimal and provides an error bound that sheds light upon how could the authors improve VP re-identification.
- Scoot: A Perceptual Metric for Facial Sketches52
The results suggest that “spatial structure” and “co-occurrence texture” are two generally applicable perceptual features in face sketch synthesis, and the first largest scale human-perception-based sketch database that can evaluate how well a metric consistent with human perception is evaluated.
- 51
- SaliencyRank: Two-stage manifold ranking for salient object detection46
This work proposes a novel graph based ranking method to detect and segment the most salient object in a scene according to its relationship to image border (background) regions, i.e., the background feature.
- 44
- Sequential Optimization for Efficient High-Quality Object Proposal Generation44
A novel object proposal algorithm, BING++, is proposed, which inherits the virtue of good computational efficiency of BING but significantly improves its proposal localization quality.
- Low-Resolution Self-Attention for Semantic Segmentation43
The Low-Resolution Self-Attention (LRSA) mechanism to capture global context at a significantly reduced computational cost, i.e., FLOPs is introduced and the effectiveness of the approach is demonstrated by building the LRFormer, a vision transformer with an encoder-decoder structure.
- 42
- Enhancing Information Maximization With Distance-Aware Contrastive Learning for Source-Free Cross-Domain Few-Shot Learning41
This paper explores a Source-Free CDFSL (SF-CDFSL) problem, in which CDFSL is addressed through the use of existing pretrained models instead of training a model with source data, avoiding accessing source data.
- EditAnything: Empowering Unparalleled Flexibility in Image Editing and Generation41
This work presents EditAnything, a novel approach that empowers users with unparalleled flexibility in editing and generating image content, and introduces an array of advanced features, including cross-image dragging, region-interactive editing, controllable layout generation, and virtual character replacement.
- Generative Multi-modal Models are Good Class-Incremental Learners40
This work proposes a novel generative multi-modal model (GMM) framework for class-incremental learning and improves by at least 14% accuracy over all the current state-of-the-art methods with significantly less forgetting in the Few-shot CIL setting.
- 40
- SRFormerV2: Taking a Closer Look at Permuted Self-Attention for Image Super-Resolution39
This paper presents SRFormer, a simple but novel method that can enjoy the benefit of large window self-attention but introduces even less computational burden, and attempts to scale up the model by further enlarging the window size and channel numbers to explore the potential of Transformer-based models.
- Revisiting Computer-Aided Tuberculosis Diagnosis39
A large-scale dataset, namely the Tuberculosis X-ray (TBX11 K) dataset, is established, which contains 11 200 chest X-ray (CXR) images with corresponding bounding box annotations for TB areas and a strong baseline, SymFormer, is proposed, for simultaneous CXR image classification and TB infection area detection.
- Bottom-Up Top-Down Cues for Weakly-Supervised Semantic Segmentation39
This work considers the task of learning a classifier for semantic segmentation using weak supervision in the form of image labels specifying objects present in the image, and adopts an Expectation-Maximization (EM) based approach, which outperforms the state-of-the-art method with a very high margin.
- 38
- Referring Camouflaged Object Detection37
This work develops a simple but strong dual-branch framework, dubbed R2CNet, with a reference branch embedding the common representations of target objects from referring images and a segmentation branch identifying and segmenting camouflaged objects under the guidance of the common representations.
- 37
- SM3Det: A Unified Model for Multi-Modal Remote Sensing Object Detection35
This paper introduces a new task called Multi-Modal Datasets and Multi-Task Object Detection (M2Det) for remote sensing, designed to accurately detect horizontal or oriented objects from any sensor modality, and proposes a unified model, SM3Det (Single Model for Multi-Modal datasets and Multi-Task object Detection).
- GET: Unlocking the Multi-modal Potential of CLIP for Generalized Category Discovery34
A Text Embedding Synthesizer (TES) is proposed to generate pseudo text embeddings for unlabelled samples, which unlocks the multi-modal potentials of CLIP and outperforms the baseline methods by a large margin on all GCD benchmarks, achieving new state-of-the-art results.
- D2ANet: Difference-aware attention network for multi-level change detection from satellite imagery34
A difference-aware attention network (D2ANet) for simultaneous building localization and multi-level change detection from the dual-temporal satellite imagery and design a difference-attention module to exploit local correlations among the multi- level changes, which improves the ability to identify damage on different scales.
- 34
- Cascade-CLIP: Cascaded Vision-Language Embeddings Alignment for Zero-Shot Semantic Segmentation33
This work introduces a series of independent decoders to align the multi-level visual features with the text embeddings in a cascaded way, forming a novel but simple framework named Cascade-CLIP that achieves superior zero-shot performance on segmentation benchmarks.
- Advancing Textual Prompt Learning with Anchored Attributes32
This work introduces an Attribute-anchored Textual Prompt learning method for vision-language models, named ATPrompt, which expands the learning space of soft prompts from the original one-dimensional category level into the multi-dimensional attribute level by incorporating multiple attribute tokens into the learnable soft prompts.
- CubemapSLAM: A Piecewise-Pinhole Monocular Fisheye SLAM System32
This work proposes a novel SLAM system with the cubemap model that utilizes the full FoV without introducing distortion from the fisheye lens, which greatly benefits the feature matching pipeline.
- Sora Generates Videos with Stunning Geometrical Consistency31
This paper introduces a new benchmark that assesses the quality of the generated videos based on their adherence to real-world physics principles, and employs a method that transforms the generated videos into 3D models, leveraging the premise that the accuracy of 3D reconstruction is heavily contingent on the video quality.
- Looking Through the Glass: Neural Surface Reconstruction Against High Specular Reflections30
A novel surface reconstruction framework, NeuS-HSR, based on implicit neural rendering, which outperforms state-of-the-art approaches for accurate and robust target surface reconstruction against HSR and designs a novel auxiliary plane module by combining physical assumptions and neural networks to generate the auxiliary plane appearance.
- FMNet: Frequency-Aware Modulation Network for SDR-to-HDR Translation29
A frequency-aware modulation block that can dynamically modulate the features according to its frequency-domain responses is designed to enhance the contrast in a frequency-adaptive way for SDR-to-HDR translation and reduce the structural distortions and artifacts in the translated low-frequency regions.
- Joint salient object detection and existence prediction29
This paper proposes a supervised learning approach for jointly addressing the salient object detection and existence prediction problems and adopts the structural SVM framework, which forms the two problems jointly in a single integrated objective function.
- 28
- 27
- Three Birds One Stone: A Unified Framework for Salient Object Segmentation, Edge Detection and Skeleton Extraction.27
This paper introduces a horizontal cascade, each component of which is densely connected to the outputs of previous component, which allows for effectively exploiting features across different levels hierarchically to effectively address the multiple pixel-wise binary regression tasks.
- 27
- 26
- Class Incremental Learning with Pre-trained Vision-Language Models26
This paper proposes an approach to exploiting pre-trained vision-language models (e.g. CLIP) that enables further adaptation instead of only using zero-shot learning of new tasks, and proposes a method for parameter retention in the adapter layers that uses a measure of parameter importance to better maintain stability and plasticity during incremental learning.
- Historical Context-based Style Classification of Painting Images via Label Distribution Learning26
This work introduces a novel knowledge distilling strategy to assist visual feature learning in the convolutional neural network for painting style classification and demonstrates the superiority of the proposed method over the state-of-the-art approaches on Painting91, OilPainting, and Pandora datasets.
- Rapid Salient Object Detection With Difference Convolutional Neural Networks25
An efficient network design is proposed that combines traditional wisdom on SOD and the representation power of modern CNNs, and a difference convolution reparameterization (DCR) strategy that embeds PDCs into standard convolutions, eliminating computation and parameters at inference.
- Masked Autoencoders are Efficient Class Incremental Learners25
This work proposes to use Masked Autoencoders (MAEs) as efficient learners for CIL and proposes a bilateral MAE framework to learn from image-level and embedding-level fusion, which produces better-quality reconstructed images and more stable representations.
- 25
- SERE: Exploring Feature Self-Relation for Self-Supervised Transformer24
This work utilizes the feature SElf-RElation (SERE) for training self-supervised ViT and enhances the relation modeling ability of ViT, resulting in stronger representations that stably improve performance on multiple downstream tasks.
- 24
- 24
- 23
- Object Proposals Estimation in Depth Image Using Compact 3D Shape Manifolds22
This work investigates the task of finding particular object shapes from a single depth image by exploiting the inherently low dimensionality in the object shape variations, which is discovered and encoded as a compact shape space.
- A Vision for Auto Research with LLM Agents21
Preliminary explorations demonstrate the feasibility and potential of Auto Research as a promising paradigm for self-improving, AI-driven research processes.
- BING++: A Fast High Quality Object Proposal Generator at 100fps21
A novel object proposal algorithm is proposed which inherits the good computational efficiency of BING but significantly improves its proposal localization quality and also recursively improves BING++'s proposals by exploiting the fact that edges in images are typically associated with object boundaries.
- TAR3D: Creating High-Quality 3D Assets Via Next-Part Prediction20
TAR3D is presented, a novel framework that consists of a 3D-aware Vector Quantized-Variational AutoEncoder and a Generative Pre-trained Transformer to generate high-quality 3D assets to migrate the multimodal unification and promising learning capabilities of the next-token prediction paradigm to conditional 3D object generation.
- TeMO: Towards Text-Driven 3D Stylization for Multi-Object Meshes20
A novel framework, dubbed TeMO, to parse multi-object 3D scenes and edit their styles under the contrast supervision at multiple levels and develops a Cross-Grained Contrast (CGC) supervision system, where a fine-grained loss between the words in the textual description and the randomly rendered images are constructed to complement the coarse-grained loss.
- 20
- 19
- Deeply Explain CNN Via Hierarchical Decomposition19
A hierarchical decomposition framework to explain CNN’s decision-making process in a top-down manner with a gradient-based activation propagation module that can decompose any intermediate CNN decision to its lower layers and find the supporting features.
- Restore Globally, Refine Locally: A Mask-Guided Scheme to Accelerate Super-Resolution Networks19
A Mask Guided Acceleration (MGA) scheme to reduce the computational costs of existing SR networks while maintaining their SR capability and experiments demonstrate that the MGA scheme reduces the FLOPs of five popular SR networks by 10% ∼ 48% with comparable or even better SR performance.
- A Glimpse to Compress: Dynamic Visual Token Pruning for Large Vision-Language Models18
A dynamic pruning framework inspired by human cognition that takes a data-driven “glimpse” and prunes irrelevant visual tokens in a single forward pass before answer generation and reduces the time and memory costs of fine-tuning with negligible performance loss.
- Towards RAW Object Detection in Diverse Conditions18
This work introduces the AODRaw dataset, which offers 7,785 high-resolution real RAW images with 135,601 annotated instances spanning 62 categories, capturing a broad range of indoor and outdoor scenes under 9 distinct light and weather conditions, and provides a comprehensive benchmark for evaluating current detection methods.
- AR-1-to-3: Single Image to Consistent 3D Object via Next-View Prediction17
AR-1-to-3 is proposed, a novel next-view prediction paradigm based on diffusion models that first generates views close to the input views, which are then utilized as contextual information to progressively synthesize farther views, and a stacked local feature encoding strategy and an LSTMbased global feature encoding strategy.
- 17
- Refinedbox: Refining for fewer and high-quality object proposals17
A computationally lightweight neural network is designed to reduce the number of proposals in object proposal generation, and can achieve state-of-the-art performance with a few proposals compared with some well-known proposal generation methods.
- 17
- 17
- KAC: Kolmogorov-Arnold Classifier for Continual Learning16
This paper introduces the Kolmogorov-Arnold Classifier (KAC), a novel classifier developed for continual learning based on the KAN structure, and explores the impact of KAN’s spline functions and introduces Radial Basis functions for improved compatibility with continual learning.
- Task-Adaptive Saliency Guidance for Exemplar-Free Class Incremental Learning16
This work introduces task-adaptive saliency for EFCIL and proposes a new framework, which it is called Task-Adaptive Saliency Supervision (TASS), for mitigating the negative effects of saliency drift between different tasks.
- A Decoupled Spatio-Temporal Framework for Skeleton-based Action Segmentation16
Joint-decoupled temporal modeling is proposed, which employs independent trainable weights to capture distinctive temporal features of each joint, inspired by the fact that different joints move at different speeds.
- CMAE-V: Contrastive Masked Autoencoders for Video Action Recognition16
This work shows that CMAE also trivially generalizes well on video action recognition without modifying the architecture and the loss criterion, by directly replacing the original pixel shift with the temporal shift.
- 16
- InterLCM: Low-Quality Images as Intermediate States of Latent Consistency Models for Effective Blind Face Restoration15
Treating low-quality images as the intermediate state of LCM, InterLCM achieves a balance between fidelity and quality by starting from earlier LCM steps and incorporates a Visual Module to extract visual features and a Spatial Encoder to capture spatial details, enhancing the fidelity of restored images.
- 15
- Image Formation Model Guided Deep Image Super-Resolution15
This work presents a simple and effective image super-resolution algorithm that imposes an image formation constraint on the deep neural networks via pixel substitution and shows that the proposed algorithm performs favorably against state-of-the-art methods.
- LSANet: Feature Learning on Point Sets by Local Spatial Attention15
This work designs a novel Local Spatial Attention (LSA) module to adaptively generate attention maps according to the spatial distribution of local regions and proposes the Spatial Feature Extractor (SFE), which constructs a branch architecture, to aggregate the spatial information with associated features in each layer of the network better.
- 15
- 15
- RPCANet$^{++}$: Deep Interpretable Robust PCA for Sparse Object Segmentation14
This work proposes RPCANet++, a sparse object segmentation framework that fuses the interpretability of RPCA with efficient deep architectures and achieves state-of-the-art performance on Infrared Small Target Detection and Vessel Segmentation tasks, while maintaining competitive results on Defect Detection tasks involving larger targets.
- Revisiting Efficient Semantic Segmentation: Learning Offsets for Better Spatial and Class Feature Alignment14
This work proposes a coupled dual-branch offset learning paradigm that explicitly learns feature and class offsets to dynamically refine both class representations and spatial image features and constructs an efficient semantic segmentation network, OffSeg.
- Dual Pyramid Generative Adversarial Networks for Semantic Image Synthesis14
A Dual Pyramid Generative Adversarial Network (DP-GAN) is proposed that learns the conditioning of spatially-adaptive normalization blocks at all scales jointly, such that scale information is bi-directionally used, and it unifies supervision at different scales.
- 14
- Mining Pixels: Weakly Supervised Semantic Segmentation Using Image Labels13
This work considers the task of learning a classifier for semantic segmentation using weak supervision, in this case, image labels specifying the objects within the image, and proposes minimizing the combination of the standard softmax loss and the KL divergence between the true latent posterior and the likelihood given by the CNN.
- Dual Frequency Transformer for Efficient SDR-to-HDR Translation12
This paper is among the first to investigate the performance of Transformers for SDR-to-HDR translation and proposes a dual frequency attention (DFA), which leverages the self-attention mechanism to separately encode the low-frequency structural information and high-frequency detail information.
- Zone Evaluation: Revealing Spatial Bias in Object Detection12
A new zone evaluation protocol is presented, extending from the traditional evaluation to a more generalized one, which measures the detection performance over zones, yielding a series of Zone Precisions (ZPs), and finds that the key lies in the human-imperceptible divergence in data patterns between objects in different zones, thus eventually forming a visible performance gap between the zones.
- RDNeRF: relative depth guided NeRF for dense free view synthesis12
This paper presents a learning-based approach named relative depth guided NeRF (RDNeRF), which jointly renders RGB images and recovers scene geometry in dense free views and recovers more accurate geometry than basic models.
- Unbiased Region-Language Alignment for Open-Vocabulary Dense Prediction11
DenseVLM is proposed, a framework designed to learn unbiased region-language alignment from powerful pretrained VLM representations that can directly replace the original VLM in open-vocabulary object detection and image segmentation methods, leading to notable performance improvements.
- 11
- Towards Sustainable Self-supervised Learning11
A Target-Enhanced Conditional (TEC) scheme which introduces two components to the existing mask-reconstruction based SSL, and proposes patch-relation enhanced targets which enhances the target given by base model and encourages the new model to learn semantic-relation knowledge from the base model by using incomplete inputs.
- Understanding Image Impressiveness Inspired by Instantaneous Human Perceptual Cues11
A novel image property is proposed, called impressiveness, that measures how images impress people with a short-term contact, based on an impression-driven model inspired by a number of important human perceptual cues.
- 11
- 11
- Anchor Token Matching: Implicit Structure Locking for Training-Free AR Image Editing10
This work introduces Implicit Structure Locking (ISLock), the first training-free editing strategy for AR visual models that achieves high-quality, structure-consistent edits without additional training and is superior or comparable to conventional editing techniques.
- 10
- 10
- 10
- 10
- Towards Universal Video MLLMs with Attribute-Structured and Quality-Verified Instructions9
Experiments across seven benchmarks show that ASID-Captioner improves fine-grained caption quality while reducing hallucinations and improving instruction following and achieves state-of-the-art performance among open-source models and is competitive with Gemini-3-Pro.
- Step-Wise Distribution-Aligned Style Prompt Tuning for Source-Free Cross-Domain Few-Shot Learning9
This paper investigates the source-free CDFSL (SF-CDFSL) problem to solve the few-shot learning (FSL) task in target domain using only a pre-trained model and a few target samples, without requiring source data or training strategies.
- 9
- Personalized Image Semantic Segmentation9
By observing the correlation among a user’s personalized images, this paper proposes a baseline method that incorporates the inter-image context when segmenting certain images and shows that this method outperforms the existing methods on the proposed dataset.
- MS-NeRF: Multi-Space Neural Radiance Fields8
This work proposes a multi-space neural radiance field (MS-NeRF) that represents the scene using a group of feature fields in parallel sub-spaces, which leads to a better understanding of the neural network toward the existence of reflective and refractive objects.
- 8
- DISTA-Net: Dynamic Closely-Spaced Infrared Small Target Unmixing8
The Dynamic Iterative Shrinkage Thresholding Network (DISTA-Net), which reconceptualizes traditional sparse reconstruction within a dynamic framework, and is the first deep learning model designed specifically for the unmixing of closely-spaced infrared small targets, achieving superior sub-pixel detection accuracy.
- Camouflaged Object Detection with Adaptive Partition and Background Retrieval8
This paper presents a novel learning framework for camouflaged object detection, called AdaptCOD, which decouples the detection process into three parts, namely localization, segmentation, and retrieval.
- NeuSmoke: Efficient Smoke Reconstruction and View Synthesis with Neural Transportation Fields8
NeuSmoke is introduced, an efficient framework for dynamic smoke reconstruction using neural transportation fields, enabling high-quality density reconstruction and novel-view synthesis from multi-view videos and achieves over 10 times faster than previous physics informed approaches.
- 8
- Ret3D: Rethinking Object Relations for Efficient 3D Object Detection in Driving Scenes8
This work introduces a simple, efficient, and effective two-stage detector, termed as Ret3D, which utilization of novel intra-frame and inter-frame relation modules to capture the spatial and temporal relations accordingly.
- 8
- A Reverse Causal Framework to Mitigate Spurious Correlations for Debiasing Scene Graph Generation7
The reconstructed causal paradigm is reconstructed into a reverse causal structure, wherein the classifier’s inputs are treated as the confounder, and both the detector’s inputs and the final predictions are viewed as causal variables, which can mitigate the spurious correlations inherent in the SGG framework, subsequently eliminating the induced biases.
- Delving Deeper into Data Scaling in Masked Image Modeling7
The study reveals that MIM can be viewed as an effective method to improve the model capacity when the scale of the training data is relatively small and MIM pre- training is data-agnostic under most scenarios, which means that the strategy of sampling pre-training data is non-critical.
- 7
- 7
- 6
- NAIPv2: Debiased Pairwise Learning for Efficient Paper Quality Estimation6
NAIPv2 is presented, a debiased and scalable framework for automated paper quality estimation and demonstrates strong generalization, with predicted scores increasing consistently across decision categories from Rejected to Oral.
- Re-Aligning Language to Visual Objects with an Agentic Workflow6
This work proposes an agentic workflow controlled by an LLM to re-align language to visual objects via adaptively adjusting image and text prompts and reveals a potential to preserve data quality along with scaling up data quantity, which further improves LOD performance from a data-alignment perspective.
- Predictive Sample Assignment for Semantically Coherent Out-of-Distribution Detection6
A concise SCOOD framework based on predictive sample assignment (PSA) that includes a dual-threshold ternary sample assignment strategy based on the predictive energy score that can significantly improve the purity of the selected ID and OOD sample sets and introduces a retraining strategy to help the model fully fit the selected auxiliary ID/OOD samples.
- Traffic Scene Parsing Through the TSP6K Dataset6
A detail refining decoder is proposed for scene parsing, which recovers the details of different semantic regions in traffic scenes owing to the proposed TSP6K dataset, and experiments show its effectiveness in parsing the traffic monitoring scenes.
- Geometry-Aware ICP for Scene Reconstruction from RGB-D Camera6
A novel ICP algorithm is proposed that provides significantly stabler registration estimation for simultaneous localization and mapping (SLAM) tasks on RGB-D camera inputs, and is readily integratable with previous ICP algorithms.
- 6
- Automatic Model Selection in Subspace Clustering via Triplet Relationships6
This paper proposes to simultaneously estimate K and segment the samples according to the local similarity relationships derived from the affinity matrix and defines a novel data structure termed the Triplet, each of which reflects a high relevance and locality among three samples which are aimed to be segmented into the same subspace.
- 6
- 6
- 6
- Direct 3D-Aware Object Insertion via Decomposed Visual Proxies5
DI (Decomposed Injection for Reference Composition and Target-integration), a novel framework that integrates interactive pose manipulation with high-fidelity 2D image synthesis to enable pose-controllable object insertion and outperforms previous methods in both geometric controllability and visual quality.
- 5
- 5
- 5
- 5
- 5
- 5
- Computer Vision5
The idea is to randomize attribute selection and no longer look for globally optimal split, but instead randomly use subset of K attributes on which to base the split.
- PhotoSketch: Internet Image Montage5
A system that composes a realistic picture from a simple freehand sketch annotated with text labels, able to automatically select suitable photographs to generate a high quality composition, using a filtering scheme to exclude undesirable images.
- Multimodal Large Models Driven SAR Image Captioning: A Benchmark Dataset and Baselines4
This article designs an annotation protocol that spans three semantic layers—object, relation, and scene—and develops a semi-automated annotation framework assisted by the large model GPT-4 V, significantly reducing human annotation cost while ensuring annotation quality.
- Enhancing Representations Through Heterogeneous Self-Supervised Learning4
The proposed Heterogeneous Self-Supervised Learning (HSSL), which enforces a base model to learn from an auxiliary head whose architecture is heterogeneous from the base model, is compatible with various self-supervised methods and achieves superior performances on various downstream tasks.
- A Simple Detector with Frame Dynamics is a Strong Tracker4
This work proposes a simple yet effective infrared tiny-object tracker that enhances tracking performance by integrating global detection and motion-aware learning with temporal priors, and proposes a trajectory constraint filtering strategy in the post-processing stage to suppress false positives and enhance tracking robustness.
- ControlSR: Taming Diffusion Models for Consistent Real-World Image Super Resolution4
ClearSR is presented, a new method that can better take advantage of latent low-resolution image (LR) embeddings for diffusion-based real-world image super-resolution (Real-ISR) and can generate more consistent SR results with LR images than existing methods.
- A Literature Review of Literature Reviews in Pattern Analysis and Machine Intelligence4
This work constructs RiPAMI, a large-scale database containing more than 3,000 review articles, and combines narrative synthesis with statistical analysis to capture structural and content-level features of PAMI reviews.
- 4
- 4
- 4
- 4
- 4
- A Comprehensive Survey on Video Scene Parsing:Advances, Challenges, and Prospects3
This survey reviews recent progress in VSP across five tasks, spanning Video Semantic Segmentation (VSS), Video Instance Segmentation (VIS), Video Panoptic Segmentation (VPS), Video Tracking \&Segmentation (VTS), and Open-Vocabulary Video Segmentation (OVVS), and traces how each family models temporal context, preserves identity, and balances accuracy against efficiency.
- A survey of historical learning: learning models with learning history3
This survey is the first survey that systematically studies the methodologies which make use of various historical statistics when training deep neural networks and encourages the community to pay attention to the think of historical learning principles when designing algorithms.
- Multi-Token Enhancing for Vision Representation Learning3
Multi-Token Enhancing (MTE) is introduced that extracts multiple auxiliary tokens simultaneously from a single model to enhance representation learning, while incurring minimal additional training costs and no additional inference costs.
- KnifeCut: Refining Thin Part Segmentation with Cutting Lines3
This work proposes KnifeCut, which is the first method to solve interactive thin structure refinement pertinently and offers the users two results, where one only focuses on the target thin part and the other provides the refinements for all thin parts that share similar features with the target one.
- Interactive Style Transfer: All is Your Palette3
This work proposes a drawing-like interactive style transfer (IST) method, by which users can interactively create a harmonious-style image, which expands the creative dimension of NST.
- 3
- 3
- 3
- 3
- 3
- WOW-Seg: A Word-free Open World Segmentation Model2
WOW-Seg is proposed, a Word-free Open World Segmentation model for segmenting and recognizing objects from open-set categories that introduces a novel visual prompt module, Mask2Token, which transforms image masks into visual tokens and ensures their alignment with the VLLM feature space.
- 2
- 2
- Benchmarking Drag⋆ for eye direction transformation and beyond2
This work systematically investigates the capability of Drag ⋆ for eye direction transformation and proposes the edge-localized point selector and zero-latent source region replacement, which can alleviate the identified limitations.
- Rendering real-world unbounded scenes with cars by learning positional bias2
A novel module to learn the positional bias between sampled points along a camera ray and target points along the incident light by multi-layer perceptrons to reconstitute the input points and view direction with regularization constraints for physical rendering.
- 2
- 2
- 2
- 2
- Mixture of Style Experts for Diverse Image Stylization1
StyleExpert is introduced, a semantic-aware framework based on the Mixture of Experts (MoE), which employs a unified style encoder, trained on the large-scale dataset of content-style-stylized triplets, to embed diverse styles into a consistent latent space.
- Sharpness-Aware Dynamic Anchor Selection for Generalized Category Discovery1
A novel method, which contains two modules: Loss Sharpness Penalty (LSP) and Dynamic Anchor Selection (DAS), which enhances the robustness of model parameters to small perturbations by minimizing the worst-case loss sharpness of the model, thereby reducing overfitting of noise samples and improving the quality of pseudo-labels.
- 1
- Visibility-Uncertainty-guided 3D Gaussian Inpainting via Scene Conceptional Learning1
A method that measures the visibility uncertainties of 3D points across different input views and uses them to guide 3DGI in utilizing complementary visual cues is proposed, which demonstrates the superior performance of the method over state-of-the-art techniques.
- Could Describe Anything Model Understand Remote Sensing Objects?1
The first systematic evaluation of the describe anything model (DAM) for localized captioning in remote sensing is presented, revealing that DAM performs robustly on visible imagery with fine-grained prompts, but exhibits significant performance degradation in infrared and SAR domains, where modality-specific distortions hinder effective spatial grounding.
- SLAN: Self-Locator Aided Network for Cross-Modal Understanding1
Self-Locator Aided Network (SLAN) is proposed for cross-modal understanding tasks without any extra gold data and demonstrates strong zero-shot and fine-tuned transferability to two localization tasks.
- 1
- 1
- Image Matching Benchmark1
This work presents the first application-oriented image matching benchmark to facilitate the analysis of matching algorithms in application level and demonstrates the effectiveness of a simple technique which is readily pluggable into any matching system to improve performance.
- Editorial Special issue on the fifth Computational Visual Media conference (CVM 2017)1
This special issue of Graphical Models includes a selection of oral papers presented at the recent Computational Visual Media conference held at Nankai University, Tianjin, China, and a novel rigidity controllable mesh deformation method, which can flexibly adjust shape rigidity by expanding the local neighborhood around a vertex beyond the first ring.
- –
- –
- –
- RoadGIE: Towards A Global-Scale Aerial Benchmark for Generalizable Interactive Road Extraction–
To improve structural consistency and mitigate performance degradation during iterative interactions, RoadGIE integrates an expert-guided prompting strategy and adapts the skeleton-based recall loss for interactive scenarios, achieving state-of-the-art performance in both segmentation accuracy and topological consistency on WorldRoadSeg-360K and other benchmarks, while maintaining efficient operation with only 3.7M parameters.
- –
- –
- Towards Understanding Token Evolution in Continual Learning of MLLMs–
This work revisits catastrophic forgetting from an interpretability perspective and introduces correctness-guided layer alignment (CLA), which allows the new model to inherit multi-layer semantic consistency from its predecessor while learning from ground-truth supervision.
- –
- Weaving Light and Time: Unified Harmonic-Geometric Representation Learning for Dense RGB-Event Parsing–
To achieve profound modal synergy, Evita explicitly embeds a suite of intrinsic co-learning modules directly into every encoder layer, and features Geometric Parallax Rectification for adaptive spatial alignment and Harmonic Spectral Resonance for texture transfer exclusively in the complex frequency domain.
- –
- AngleRoCL: Angle-Robust Concept Learning for Physically View-Invariant T2I Adversarial Patches–
The angle robustness of T2I adversarial patches is studied comprehensively, revealing their angle-robust issues, and AngleRoCL is introduced, a simple and flexible approach that significantly enhances the angle robustness of T2I adversarial patches compared to baseline methods.
- –
- –
- Revisiting Data Challenges of Computational Pathology: A Pack-based Multiple Instance Learning Training Framework–
This work proposes a pack-based MIL framework, which packs multiple sampled, variable-length feature sequences into fixed-length ones, enabling batched training while preserving data heterogeneity, and introduces a residual branch that composes discarded features from multiple slides into a hyperslide which is trained with tailored labels.
- –
- –
- –
- –
- –
- PR-MIM: Delving Deeper into Partial Reconstruction in Masked Image Modeling–
This work proposes a progressive reconstruction strategy and a furthest sampling strategy to reconstruct those thrown tokens in an extremely lightweight way instead of completely abandoning them and validates the effectiveness of the proposed method across various existing frameworks.
- Towards Spatial Equilibrium Object Detection–
This paper proposes a spatial equilibrium label assignment (SELA) to alleviate the spatial disequilibrium problem by injecting the prior spatial weight into the optimization process of detectors and challenges the conventional sense of object detectors and shows the indispensability of spatial equilibrium.
- –
- –
- –
- Model-Agnostic Structured Sparsification with Learnable Channel Shuffle.–
A model-agnostic structured sparsification method for efficient network compression that automatically induces structurally sparse representations of the convolutional weights, thereby facilitating the implementation of the compressed model with the highly-optimized group convolution.
- –
- –
- Pedestrian Detection with a Directly-Cascaded Deconvolution-Convolution Structure–
A novel neural network structure which can be flexibly combined with powerful object detection systems for boosting pedestrian detection and incorporates the features learned in different layers in a novel and unified fashion is proposed.
- –
- –
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.