Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Shuicheng YanSuggested from co-authorship
- Wei LiuSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works94 from public data
- Revisiting Dilated Convolution: A Simple Approach for Weakly- and Semi-Supervised Semantic Segmentation627
It is found that varying dilation rates can effectively enlarge the receptive fields of convolutional kernels and more importantly transfer the surrounding discriminative information to non-discriminative object regions, promoting the emergence of these regions in the object localization maps.
- A Sufficient Condition for Convergences of Adam and RMSProp470
An alternative easy-to-check sufficient condition is introduced, which merely depends on the parameters of the base learning rate and combinations of historical second-order moments, to guarantee the global convergence of generic Adam/RMSProp for solving large-scale non-convex stochastic optimization.
- Evaluation of Combined Artificial Intelligence and Radiologist Assessment to Interpret Screening Mammograms409
This diagnostic accuracy study evaluates whether artificial intelligence can overcome human mammography interpretation limits with a rigorous, unbiased evaluation of machine learning algorithms.
- Central Similarity Quantization for Efficient Image and Video Retrieval392
The Central Similarity Quantization (CSQ) is proposed, a new similarity metric with which the hash codes of similar data pairs are encouraged to approach a common center and those for dissimilar pairs to converge to different centers, to improve hash learning efficiency and retrieval accuracy.
- Temporally Grounding Natural Sentence in Video345
An effective and efficient method that grounds (i.e., localizes) natural sentences in long, untrimmed video sequences using a novel Temporal GroundNet to temporally capture the evolving fine-grained frame-by-word interactions between video and sentence.
- PromptDet: Towards Open-Vocabulary Detection Using Uncurated Images245
The goal of this work is to establish a scalable pipeline for expanding an object detector towards novel/unseen categories, using zero manual annotations, and the proposed detector, termed as PromptDet, shows superior performance over existing approaches with fewer additional training images andzero manual annotations whatsoever.
- Joint Task-Recursive Learning for Semantic Segmentation and Depth Estimation237
A novel joint Task-Recursive Learning (TRL) framework for the closing-loop semantic segmentation and monocular depth estimation tasks and propagate previous learning experiences on both tasks into the next network evolution by explicitly concatenating previous responses.
- Deep Self-Taught Learning for Weakly Supervised Object Localization207
A deep self-taught learning approach is proposed, which makes the detector learn the object-level features reliable for acquiring tight positive samples and afterwards re-train itself based on them, which progressively improves its detection ability and localizes more informative positive samples.
- NMS by Representative Region: Towards Crowded Pedestrian Detection by Proposal Pairing197
A novel Representative Region NMS (R2NMS) approach leveraging the less occluded visible parts, effectively removing the redundant boxes without bringing in many false positives is proposed.
- Multi-View Image Generation from a Single-View193
This paper proposes a novel image generation model termed VariGANs, which combines the merits of the variational inference and the Generative Adversarial Networks (GANs), and generates the target image in a coarse-to-fine manner instead of a single pass which suffers from severe artifacts.
- Video-Based Person Re-Identification With Accumulative Motion Context181
The experimental results demonstrate that the proposed AMOC network outperforms state-of-the-arts for video-based re-identification significantly and confirm the advantage of exploiting long-range motion context forVideo-based person re-Identification, validating the motivation evidently.
- Policy Optimization with Demonstrations180
It is shown that POfD induces implicit dynamic reward shaping and brings provable benefits for policy improvement, and can be combined with policy gradient methods to produce state-of-the-art results, as demonstrated experimentally on a range of popular benchmark sparse-reward tasks.
- Neural Person Search Machines171
Evaluations on two benchmark datasets have demonstrated that the Neural Person Search Machines (NPSM) developed can outperform current state-of-the-arts in both mAP and top-1 evaluation protocols.
- MSMDFusion: Fusing LiDAR and Camera at Multiple Scales with Multi-Depth Seeds for 3D Object Detection152
A novel framework with better utilization of the depth information and fine-grained cross-modal interaction between LiDAR and camera is proposed, which consists of two important components and achieves state-of-the-art results on both 3D object detection and tracking tasks without using test-time-augmentation and ensemble techniques.
- 119
- Video Scene Parsing with Predictive Feature Learning117
It is experimentally proved that the learned predictive features in the model are able to significantly enhance the video parsing performance by combining with the standard image parsing network.
- MTL-NAS: Task-Agnostic Neural Architecture Search Towards General-Purpose Multi-Task Learning113
A novel single-shot gradient-based search algorithm that closes the performance gap between the searched architectures and the final evaluation architecture and achieves a single model, which can be directly used for evaluation without (re-)training from scratch.
- LLaVA-MoLE: Sparse Mixture of LoRA Experts for Mitigating Data Conflicts in Instruction Finetuning MLLMs107
Extensive experiments proved that LLaVA-MoLE effectively mitigates the data conflict issue when mixing multiple distinct instruction datasets with various configurations, and achieves consistent performance gains over the strong plain-LoRA baselines.
- Localizing Natural Language in Videos107
A localizing network (LNet), working in an end-to-end fashion, to tackle the NLVL task, which first match the natural sentence and video sequence by cross-gated attended recurrent networks to exploit their fine-grained interactions and generate a sentence-aware video representation.
- 100
- Learning to segment with image-level annotations99
This paper explores a more challenging problem by learning to segment under image-level annotations by incorporating the hypotheses-aware classification and cross-image contextual refinement in a supervised manner, and proposes two network training strategies for achieving good segmentation performance.
- Left-Right Comparative Recurrent Model for Stereo Matching89
A novel left- right comparative recurrent model to perform left-right consistency checking jointly with disparity estimation and a soft attention mechanism is introduced, which employs the learned error maps for better guiding the model to selectively focus on refining the unreliable regions at the next recurrent step.
- Modular Generative Adversarial Networks82
ModularGAN consists of several reusable and composable modules that carry on different functions that can be trained simultaneously, leveraging data from all domains, and then combined to construct specific GAN networks at test time, according to the specific image translation task.
- TimeMarker: A Versatile Video-LLM for Long and Short Video Understanding with Superior Temporal Localization Ability81
TimeMarker is introduced, a versatile Video-LLM designed for high-quality dialogue based on video content, emphasizing temporal localization, and achieves state-of-the-art performance across multiple benchmarks, excelling in both short and long video categories.
- 71
- 62
- FoveaNet: Perspective-Aware Urban Scene Parsing61
This work proposes a new FoveaNet model to fully exploit the perspective geometry of scene images and address the common failures of generic parsing models, and introduces a new dense CRFs model that takes the Perspective geometry as a prior potential.
- Reversible Recursive Instance-Level Object Segmentation61
This work proposes a novel Reversible Recursive Instance-level Object Segmentation (R2-IOS) framework that consists of a reversible proposal refinement sub-network that predicts bounding box offsets for refining the object proposal locations, and an instance-level segmentation sub- network that generates the foreground mask of the dominant object instance in each proposal.
- 55
- InstaGen: Enhancing Object Detection by Training on Synthetic Dataset50
This enhanced version of diffusion model, termed as InstaGen, can serve as a data synthe-sizer, to enhance object detectors by training on its generated samples, demonstrating superior performance over existing state-of-the-art methods in open-vocabulary and data-sparse scenarios.
- Learning Object-Wise Semantic Representation for Detection in Remote Sensing Imagery49
An enhanced feature pyramid network is first designed to better extract hierarchical discriminative visual features to suppress background clutter as well as better estimate proposals, and specifically introduces a semantic segmentation module to guide horizontal proposals detection.
- Learning with rethinking: Recurrently improving convolutional neural networks through feedback42
By adding a feedback layer and producing the emphasis vector, the model is able to recurrently boost the performance based on previous prediction, and can be employed to boost any pre-trained models.
- Weighted AdaGrad with Unified Momentum42
AdaUSM is proposed, which has the main characteristics that it incorporates a unified momentum scheme which covers both the heavy ball momentum and the Nesterov accelerated gradient momentum, and adopts a novel weighted adaptive learning rate that can unify the learning rates of AdaGrad, AccAdaGrad, Adam, and RMSProp.
- 40
- PS-RCNN: Detecting Secondary Human Instances in a Crowd via Primary Object Suppression40
This work introduces a variant of two-stage detectors called PS-RCNN, which significantly improves recall and AP on CrowdHuman and Widerperson, and introduces a High Resolution RoI Align module to retain as much of fine-grained features of visible parts of the heavily occluded humans as possible.
- AlignSAM: Aligning Segment Anything Model to Open Context via Reinforcement Learning39
A novel framework, termed AlignSAM, designed for automatic prompting for aligning SAM to an open context through reinforcement learning, which enables the generality of the SAM model across diverse downstream tasks while keeping its parameters frozen.
- Matten: Video Generation with Mamba-Attention38
This paper introduces Matten, a cutting-edge latent diffusion model with Mamba-Attention architecture for video generation that has competitive performance with the current Transformer-based and GAN-based models in benchmark performance, achieving superior FVD scores and efficiency.
- OV-DINO: Unified Open-Vocabulary Detection with Language-Aware Selective Fusion38
A novel unified open-vocabulary detection method called OV-DINO is proposed, which is pre-trained on diverse large-scale datasets with language-aware selective fusion in a unified framework, and a Unified Data Integration pipeline is introduced to enable end-to-end training and eliminate noise from pseudo-label generation by unifying different data sources into detection-centric data format.
- Weakly Supervised Semantic Segmentation Via Progressive Patch Learning37
A novel “Progressive Patch Learning” approach is proposed to improve the local details extraction of the classification, producing the CAM better covering the whole object rather than only the most discriminative regions as in CAMs obtained in conventional classification models.
- RoboTron-Drive: All-in-One Large Multimodal Model for Autonomous Driving36
This work proposes RoboTron-Drive, a general large multimodal model designed to process diverse data inputs, such as images and multi-view videos, while performing a broad spectrum of AD tasks, including perception, prediction, and planning.
- ARMANI: Part-level Garment-Text Alignment for Unified Cross-Modal Fashion Design36
This work argues that simply aligning texture and garment information is not sufficient to capture the semantics of the visual information and therefore proposes MaskCLIP, a unified cross-modal fashion designer with part-level garment-text alignment that outperforms existing state-of-the-art cross- modal image synthesis approaches.
- 35
- 34
- Bidirectional image-sentence retrieval by local and global deep matching33
The proposed local and global deep matching model outperforms the state-of-the-art bidirectional retrieval approaches on the Flickr8K, Flickr30K, and MSCOCO datasets and the image and sentence representations exploited in local andglobal levels are demonstrated to play synergic and complementary roles for biddirectional image-sentence retrieval.
- 32
- Multi-Path Feedback Recurrent Neural Networks for Scene Parsing32
This paper considers the scene parsing problem and proposes a novel Multi-Path Feedback recurrent neural network (MPF-RNN), which can enhance the capability of RNNs in modeling long-range context information at multiple levels and better distinguish pixels that are easy to confuse.
- AeDet: Azimuth-Invariant Multi-View 3D Object Detection31
A camera-decoupled virtual depth is introduced to unify the depth prediction for the images with different camera intrinsic parameters and the resultant detector is dubbed Azimith-equivariant Detector (AeDet), topping the recent multi-view 3D object detectors such as PETRv2 and BEVDepth by a large margin.
- Two-stage Visual Cues Enhancement Network for Referring Image Segmentation31
Through the two-stage enhancement, the proposed TV-Net enjoys better performances in learning fine-grained matching behaviors between the natural language expression and image, especially when the visual information of the referent is inadequate, thus produces better segmentation results.
- 30
- Curriculum Multi-Negative Augmentation for Debiased Video Grounding29
This work proposes multiple negative augmentations in a hierarchical way, including cross-video augmentations from clip-/video-level, and self-shuffled augmentations with masks, to discourage the standard VG model's behavior of exploiting such temporal annotation biases and improve the model generalization ability.
- 29
- Delving deep into the imbalance of positive proposals in two-stage object detection23
This work incorporates an R-CNN Gradient Annealing (RGA) strategy to enhance the impact of positive proposals in the early training stage and a set of Parallel R- CNN Modules (PRM) with different positive/negative sampling ratios during training on one same backbone to tackle imbalance issues.
- 23
- 23
- 22
- 22
- Modeling Varying Camera-IMU Time Offset in Optimization-Based Visual-Inertial Odometry21
This work proposes a nonlinear optimization-based monocular visual inertial odometry (VIO) with varying camera-IMU time offset modeled as an unknown variable that is able to handle the rolling-shutter effects and imperfect sensor synchronization in a unified way.
- 19
- 18
- CLIP-GS: Unifying Vision-Language Representation with 3D Gaussian Splatting14
This paper introduces the GS Tokenizer to generate serialized gaussian tokens, which are then processed through transformer layers pre-initialized with weights from point cloud models, resulting in the 3DGS embeddings, and introduces an image voting loss to guide the directionality and convergence of gradient optimization.
- 14
- 14
- 13
- 13
- Robust Scene Classification with Cross-Level LLC Coding on CNN Features13
This framework first fine-tunes multi-level CNNs in a cascaded way, then extracts multi- level CNN features to learn a cross-level universal codebook, and finally performs locality-constrained linear coding (LLC) and max-pooling on the patches of all levels to form the final representation.
- Photo to Family Tree: Deep Kinship Understanding for Nuclear Family Photos11
This paper extends kinship recognition from two persons to a nuclear family consisting of multiple persons and proposes a novel Deep Kinship Recognition (DKR) framework, which gains superior performance on both Group-Face dataset and TSKinFace dataset, compared with state-of-the-arts.
- Suspected Objects Matter: Rethinking Model's Prediction for One-stage Visual Grounding10
A Suspected Object Transformation mechanism (SOT), which can be seamlessly integrated into existing CNN and Transformer-based one-stage visual grounders to encourage the target object selection among the suspected ones.
- Fewer Tokens and Fewer Videos: Extending Video Understanding Abilities in Large Vision-Language Models8
This paper presents a cost-effective video-LVLM that enhances model architecture, introduces innovative training strategies, and identifies the most effective types of video instruction data, and delve into the influence of video instruction data in limited-resource settings.
- VidCompress: Memory-Enhanced Temporal Compression for Video Understanding in Large Language Models8
Experiments on several VideoQA datasets and comprehensive benchmarks demonstrate that VidCompress efficiently models complex temporal-spatial relations and significantly outperforms existing Video-LLMs.
- MT-Net Submission to the Waymo 3D Detection Leaderboard8
This work designs a 2D backbone to utilize multi-scale features for better detecting objects with various sizes, together with an optimal transport-based target assignment strategy, which dynamically assigns richer supervision signals to the detection candidates.
- Image-level to Pixel-wise Labeling: From Theory to Practice8
Theoretically, the effects of image-level labels on pixel-wise segmentation from the perspective of information theory are analyzed and discussed, and in practice, an end-to-end segmentation model is built by fusing the image- level andpixel-wise labeling networks.
- Multiple Object Tracking Challenge Technical Report for Team MT_IoT5
An improved human detector is designed and associated most of detection is associated most of detection to guarantee the integrity of the motion trajectory and a location-wise matching matrix is proposed to obtain more accurate trace matching.
- Central Similarity Hashing via Hadamard matrix.5
A new hash center network (HCN) that learns hashing functions by optimizing the central similarity w.r.t.\ these hash centers is devised and can be applied for both image and video hashing.
- 4
- Weakly-Supervised 3D Visual Grounding Based on Visual Language Alignment4
This work proposes 3D-VLA, a weakly supervised approach for 3D visual grounding based on Visual Language Alignment, which exploits the superior ability of current large-scale vision-language models on aligning the semantics between texts and 2D images, as well as the naturally existing correspondences between 2D images and 3D point clouds.
- MindBench: A Comprehensive Benchmark for Mind Map Structure Recognition and Analysis4
The new benchmark named MindBench is introduced, which not only includes meticulously constructed bilingual authentic or synthetic images, detailed annotations, evaluation metrics and baseline models, but also specifically designs five types of structured understanding and parsing tasks.
- 4
- UniDiff: Advancing Vision-Language Models with Generative and Discriminative Learning3
UniDiff is a unified multi-modal model that integrates image-text contrastive learning (ITC), text-conditioned image synthesis learning (IS), and reciprocal semantic consistency modeling (RSC) and effectively learns aligned semantics and mitigates the issue of semantic collapse during fine-tuning on small datasets.
- 2
- Orthogonal Projection Subspace to Aggregate Online Prior-knowledge for Continual Test-time Adaptation1
A novel pipeline, Orthogonal Projection Subspace to aggregate online Prior-knowledge, dubbed OoPk is proposed, which first project a tuning subspace orthogonally which allows the model to adapt to new domains while preserving the knowledge integrity of the pre-trained source model to alleviate catastrophic forgetting.
- 1
- Enhancing Robustness of Vision-Language Models through Orthogonality Learning and Self-Regularization1
An orthogonal finetuning method for efficiently updating pretrained weights which enhances robustness and generalization, while a cross-regularization strategy is further exploited to maintain the stability in terms of zero-shot generalization of VLMs, dubbed OrthCR.
- MRStyle: A Unified Framework for Color Style Transfer with Multi-Modality Reference1
The proposed MRStyle framework is a comprehensive framework that enables color style transfer using multi-modality reference, including image and text, and outperforms the state-of-the-art in both qualitative and quantitative evaluations.
- Accurate On-Road Vehicle Detection with Deep Fully Convolutional Networks1
This paper presents a framework utilizing fully convolutional networks (FCN) to produce bounding boxes with high confidence to contain a vehicle, and bounding box location refinement with SVM to further improve localization accuracy.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Central Similarity Hashing for Efficient Image and Video Retrieval–
The Central Similarity Hashing (CSH) is proposed, a new similarity metric with which the hash codes for similar data pairs are encouraged to approach a common center and those for dissimilar pairs to converge to different centers to improve hash learning efficiency and retrieval accuracy.
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.