Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks83 from public data
- Exploring Cross-Image Pixel Contrast for Semantic Segmentation596
This work raises a pixel-wise metric learning paradigm for semantic segmentation, by explicitly exploring the structures of labeled pixels, which were rarely explored before, and hopes this work will encourage the community to rethink the current de facto training paradigm in semantic segmentsation.
- Rethinking Semantic Segmentation: A Prototype View377
This study uncovers several limitations of such parametric segmentation regime, and proposes a nonparametric alternative based on non-learnable prototypes, which allows the model to directly shape the pixel embedding space, by optimizing the arrangement between embedded pixels and anchored prototypes.
- A Survey on Deep Learning Technique for Video Segmentation316
This survey comprehensively review two basic lines of research — generic object segmentation in videos, and video semantic segmentation — by introducing their respective task settings, background concepts, perceived need, development history, and main challenges.
- Video Object Segmentation with Episodic Graph Memory Networks303
This work exploits an episodic memory network, organized as a fully connected graph, to store frames as nodes and capture cross-frame correlations by edges and yields a neat yet principled framework, which can generalize well both one-shot and zero-shot video object segmentation tasks.
- Visual Recognition with Deep Nearest Centroids217
DNC is devised, a conceptually elegant yet surprisingly effective network for large-scale visual recognition, by revisiting Nearest Centroids, one of the most classic and simple classifiers, and brings fundamental insights into related fields.
- MATNet: Motion-Attentive Transition Network for Zero-Shot Video Object Segmentation212
A novel end-to-end learning neural network, i.e., MATNet, for zero-shot video object segmentation (ZVOS), motivated by the human visual attention behavior, which leverages motion cues as a bottom-up signal to guide the perception of object appearance.
- Regional Semantic Contrast and Aggregation for Weakly Supervised Semantic Segmentation206
This work proposes regional semantic contrast and aggregation (RCA), equipped with a regional memory bank to store massive, diverse object patterns appearing in training data, which acts as strong support for exploration of dataset-level semantic structure.
- Deep Hierarchical Semantic Segmentation198
HSSN is devised, a general HSS framework that tackles two critical issues in this task: how to efficiently adapt existing hierarchy-agnostic segmentation networks to the HSS setting, and how to leverage the hierarchy information to regularize HSS network learning.
- Face Forensics in the Wild195
This work constructs a novel large-scale dataset, called FFIW10K, which comprises 10,000 high-quality forgery videos, with an average of three human faces in each frame, and proposes a novel algorithm to tackle the task of multi-person face forgery detection.
- 179
- 156
- Group-Wise Semantic Mining for Weakly Supervised Semantic Segmentation147
This work designs a graph neural network (GNN) for group-wise semantic segmentation (WSSS), wherein input images are represented as graph nodes, and the underlying relations between a pair of images are char- acterized by an efficient co-attention mechanism.
- Detail-Preserving Transformer for Light Field Image Super-resolution140
This paper proposes a novel formulation built upon Transformers, by treating LFSR as a sequence-to-sequence reconstruction task, and proposes a detail-preserving Transformer (termed as DPT), by leveraging gradient maps of light field to guide the sequence learning.
- Cascaded Human-Object Interaction Recognition130
This work introduces a cascade architecture for a multi-stage, coarse-to-fine HOI understanding, and makes the framework flexible to perform fine-grained pixel-wise relation segmentation; this provides a new glimpse into better relation modeling.
- Weakly Supervised 3D Object Detection from Lidar Point Cloud127
This work proposes a weakly supervised approach for 3D object detection, only requiring a small set of weakly annotated scenes, associated with a few precisely labeled object instances, achieved by a two-stage architecture design.
- Local-Global Context Aware Transformer for Language-Guided Video Segmentation120
This work presents Locater, which augments the Transformer architecture with a finite memory so as to query the entire video with the language expression in an efficient manner, and outperforms previous state-of-the-arts solutions.
- 119
- Volumetric memory network for interactive medical image segmentation109
A novel Volumetric Memory Network is proposed, dubbed as VMN, to enable segmentation of 3D medical images in an interactive manner and leads to a robust interactive segmentation engine, which can generalize well to various types of user annotations.
- 109
- 106
- 97
- Core: Cooperative Reconstruction for Multi-Agent Perception96
Core is presented, a conceptually simple, effective and communication-efficient model for multi-agent cooperative perception that addresses the task from a novel perspective of cooperative reconstruction, based on two key insights: cooperating agents together provide a more holistic observation of the environment and the holistic observation can serve as valuable supervision to explicitly guide the model learning how to reconstruct the ideal observation based on collaboration.
- CLUSTSEG: Clustering for Universal Segmentation92
These innovations closely link CLUSTSEG to EM clustering and make it a transparent and powerful framework that yields superior results across the above segmentation tasks.
- 86
- 83
- Differentiable Multi-Granularity Human Representation Learning for Instance-Aware Human Semantic Parsing78
A new bottom-up regime is proposed to learn category-level human semantic segmentation as well as multi- person pose estimation in a joint and end-to-end manner that exploits structural information over different human granularities and eases the difficulty of person partitioning.
- 77
- Web-Supervised Network with Softly Update-Drop Training for Fine-Grained Visual Classification66
A novel approach to remove irrelevant samples from the real-world web images during training, and only utilize useful images for updating the networks is proposed, which can alleviate the harmful effects caused by irrelevant noisy web images to achieve better performance.
- Locality-Aware Inter-and Intra-Video Reconstruction for Self-Supervised Correspondence Learning63
Lir is developed, a locality-aware inter-and intra-video reconstruction method that fills in three missing pieces, i.e., instance discrimination, location awareness, and spatial compactness, of self-supervised correspondence learning puzzle, and surpasses self- supervised state-of-the-arts on label propagation tasks including objects, semantic parts, and keypoints.
- Image Segmentation in Foundation Model Era: A Survey61
A fresh, comprehensive, and systematic survey of cutting-edge research centered around FM-driven image segmentation, providing insights into the emergence of segmentation knowledge from FMs like CLIP, Stable Diffusion, and DINO.
- Visual Abductive Reasoning59
A new task and dataset, Visual Abductive Reasoning (VAR), for examining abductive reasoning ability of machine intelligence in everyday visual situations, and devise a strong baseline model, REASONER (causal-and-cascaded reasoning Transformer), which surpasses many famous video-language models, while still being far behind human performance.
- FedFA: Federated Feature Augmentation57
Key to the approach is the determination of a meaningful Gaussian variance, which is accomplished by taking into account not only biased data of each individual client, but also underlying feature statistics characterized by all participating clients.
- 54
- GeneSegNet: a deep learning framework for cell segmentation by integrating gene expression and imaging49
This work developed a deep-learning-based method, GeneSegNet, that integrates both gene expression and imaging information to perform cell segmentation and employs a recursive training strategy to deal with noisy training labels.
- 47
- Target-Aware Object Discovery and Association for Unsupervised Video Multi-Object Segmentation47
This paper proposes to combine foreground region estimation and instance grouping together in one network, and additionally introduces temporal guidance for segmenting each frame, enabling more accurate object discovery and outperforms state-of theart methods both in segmentation accuracy and inference speed.
- 45
- SCLNet: Spatial context learning network for congested crowd counting45
A spatial context learning network (SCLNet) for congested crowd counting that is competitive to the other state-of-the-art methods on the crowd localization task with the UCF-QNRF dataset, and the results demonstrate the effectiveness of the model.
- 41
- Unified Mask Embedding and Correspondence Learning for Self-Supervised Video Segmentation32
A unified framework which simultaneously models cross-frame dense correspondence for locally discriminative feature learning and embeds object-level context for target-mask decoding is developed and is able to directly learn to perform mask-guided sequential segmentation from unlabeled videos.
- 31
- 28
- 26
- 21
- 21
- 19
- Quality-Aware Memory Network for Interactive Volumetric Image Segmentation19
The proposed quality-aware memory network leads to a robust interactive segmentation engine, which can generalize well to various types of user annotations and demonstrate the superiority of the approach in comparison with existing techniques.
- Multi-Granular Semantic Mining for Weakly Supervised Semantic Segmentation18
Through heterogeneous graph learning, the proposed Hgnn is able to land a comprehensive understanding of object patterns, leading to more accurate semantic concept grounding, and outperforms the current state-of-the-art approaches on the popular PASCAL VOC 2012 and COCO 2014 benchmarks.
- Label-Efficient Few-Shot Semantic Segmentation with Unsupervised Meta-Training17
The goal of this paper is to alleviate the training cost for few-shot semantic segmentation models by taking a pioneering step towards label-efficient training of FSS models from fully unlabeled training data, or additionally a few labeled samples to enhance the performance.
- 15
- 15
- Seeing What Matters: Empowering CLIP with Patch Generation-to-Selection14
This work presents a concise yet effective approach called Patch Generation-to-Selection (CLIP-PGS) to enhance CLIP’s training efficiency while preserving critical semantic con tent, setting new state-of-the-art results in zero-shot classification and retrieval tasks.
- 14
- Learning Anchor Transformations for 3D Garment Animation13
An anchor-based deformation model, namely AnchorDEF, to predict 3D garment animation from a body motion sequence that deforms a garment mesh template by a mixture of rigid transformations with extra nonlinear displacements and an adaptive anchor updating is proposed to optimize the anchor position.
- LFMamba: Light Field Image Super-Resolution with State Space Model12
This work carefully devise a basic SSM block characterized by an efficient SS2D mechanism that facilitates more effective and efficient feature learning on these 2D slices of 4D LFs, and introduces an SSM-based network for LFSR termed LFMamba.
- 12
- 12
- Abrupt motion tracking via nearest neighbor field driven stochastic sampling12
A novel sampling-based method in the Bayesian filtering framework to address the problem of abrupt motion tracking and develops an abrupt-motion detection scheme which can discover the presence of abrupt motions during online tracking.
- Decouple Before Align: Visual Disentanglement Enhances Prompt Tuning10
This paper delves into a previously overlooked information asymmetry issue in PT, where the visual modality mostly conveys more context than the object-oriented textual modality, and proposes DAPT, an effective PT framework based on an intuitive decouple-before-align concept.
- Many-to-One Distribution Learning and K-Nearest Neighbor Smoothing for Thoracic Disease Identification10
This paper proposes many-to-one distribution learning (MODL) and K-nearest neighbor smoothing (KNNS) methods from two perspectives to improve a single model's disease identification performance, rather than focusing on an ensemble of models.
- Refined video segmentation through global appearance regression8
A novel segmentation framework based on a two-stream deep convolution network that exploits the object’s robust pixel-level features within all the video frames and generates foreground likelihood maps with sufficient details to achieve accurate segmentation in unconstrained videos.
- 7
- Instance significance guided multiple instance boosting for robust visual tracking7
This paper incorporates the instance significance estimation into the online MILBoost framework, and incorporates the estimates within a new boosting procedure for more effectively selection of weak classifiers.
- Online Visual Tracking using Multiple Instance Learning with Instance Significance Estimation.6
This paper follows the online boosting framework, and proposes a new criterion for the selection of weak classifiers, and outperforms both existing MIL based and boosting based trackers.
- 5
- nsDB: Architecting the Next Generation Database by Integrating Neural and Symbolic Systems5
This paper proposes nsDB, a novel neuro-symbolic database system that integrates neural and symbolic system architectures natively to address the weaknesses of each, providing a strong database capable of data managing, model learning, and complex analytical query processing over multi-modal data.
- 5
- 4
- Prompt-Driven Continual Graph Learning2
A novel prompt-driven continual graph learning (PROMPTCGL) framework, which learns a separate prompt for each incoming task and maintains the underlying graph neural network model fixed, and develops a personalized prompt generator to generate tailored prompts for each graph node while minimizing the number of prompts needed.
- 2
- 2
- 1
- 1
- 1
- A Sparse Error Compensation Based Incremental Principal Component Analysis Method for Foreground Detection1
This paper proposes a sparse error compensation based incremental principal component analysis method, which robustly updates background subspace and estimates foreground, which reduces the interference of undesirable foreground signal.
- –
- –
- Lava : Language Driven Scalable and Versatile Traffic Video Analytics–
This work builds Lava, a system that accepts natural language queries and retrieves traffic targets across multiple levels of granularity and arbitrary categories, and develops a novel benchmark by providing diverse, semantically rich natural language predicates and fine-grained annotations for multiple videos.
- –
- –
- Automatic Foreground Seeds Discovery for Robust Video Saliency Detection–
This work aims to discover robust foreground priors as a complement to background priors so that the performance of video saliency detection in unconstrained videos can be improved.
- Spectral Context Matching for Video Object Segmentation Under Occlusion–
A new method, Spectral Context Matching (SCM), is proposed, for a coarse object location, to obtain an initial localization and utilize a spatial-temporal graphical model on the estimated object region to get an accurate segmentation.
- –
Publication data from OpenAlex; 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.