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
- Licheng JiaoSuggested from co-authorship
- Cristian SminchisescuSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works148 from public data
- 1,548
- 1,147
- Multiobjective Immune Algorithm with Nondominated Neighbor-Based Selection486
The statistical analysis based on three performance metrics show that the unique selection method is effective, and NNIA is an effective algorithm for solving multiobjective optimization problems.
- 484
- Unsupervised Feature Learning for RGB-D Based Object Recognition394
HMP uses sparse coding to learn hierarchical feature representations from raw RGB-D data in an unsupervised way and enables superior object recognition results using linear support vector machines.
- 392
- Kernel Descriptors for Visual Recognition375
This work highlights the kernel view of orientation histograms, and shows that they are equivalent to a certain type of match kernels over image patches, and designs a family of kernel descriptors which provide a unified and principled framework to turn pixel attributes into compact patch-level features.
- 348
- A Joint Model of Language and Perception for Grounded Attribute Learning324
This work presents an approach for joint learning of language and perception models for grounded attribute induction, which includes a language model based on a probabilistic categorial grammar that enables the construction of compositional meaning representations.
- Adaptive Wing Loss for Robust Face Alignment via Heatmap Regression314
A novel loss function is proposed, named Adaptive Wing loss, that is able to adapt its shape to different types of ground truth heatmap pixels, that penalizes loss more on foreground pixels while less on background pixels.
- Discriminatively Trained Sparse Code Gradients for Contour Detection291
This work shows that contour detection accuracy can be significantly improved by computing Sparse Code Gradients (SCG), which measure contrast using patch representations automatically learned through sparse coding, which is verified on the NYU Depth Dataset.
- Twin Gaussian Processes for Structured Prediction279
Twin Gaussian processes (TGP), a generic structured prediction method that uses Gaussian process priors on both covariates and responses, both multivariate, and estimates outputs by minimizing the Kullback-Leibler divergence between two GP modeled as normal distributions over finite index sets of training and testing examples, is described.
- 277
- Knowledge-Enhanced Hierarchical Graph Transformer Network for Multi-Behavior Recommendation271
A Knowledge-Enhanced Hierarchical Graph Transformer Network (KHGT) is proposed, to investigate multi-typed interactive patterns between users and items in recommender systems and consistently outperforms many state-of-the-art recommendation methods across various evaluation settings.
- Efficient Match Kernel between Sets of Features for Visual Recognition265
It is shown that bag-of-words representations commonly used in conjunction with linear classifiers can be viewed as special match kernels, which count 1 if two local features fall into the same regions partitioned by visual words and 0 otherwise.
- 263
- Traffic Flow Forecasting with Spatial-Temporal Graph Diffusion Network251
A new traffic prediction framework–Spatial-Temporal Graph Diffusion Network (ST-GDN) is developed which learns not only the local region-wise geographical dependencies, but also the spatial semantics from a global perspective, and a multi-scale attention network is developed to empower ST-G DN with the capability of capturing multi-level temporal dynamics.
- Hierarchical Matching Pursuit for Image Classification: Architecture and Fast Algorithms238
This paper proposes hierarchical matching pursuit (HMP), which builds a feature hierarchy layer-by-layer using an efficient matching pursuit encoder that includes three modules: batch (tree) orthogonal matching pursuit, spatial pyramid max pooling, and contrast normalization.
- Knowledge-aware Coupled Graph Neural Network for Social Recommendation233
A Knowledge-aware Coupled Graph Neural Network (KCGN) that jointly injects the inter-dependent knowledge across items and users into the recommendation framework and enables the high-order user- and item-wise relation encoding by exploiting the mutual information for global graph structure awareness.
- 231
- RichDreamer: A Generalizable Normal-Depth Diffusion Model for Detail Richness in Text-to-3D217
This paper proposes to learn a generalizable Normal-Depth diffusion model for 3D generation by training on the large-scale LAION dataset together with the generalizable image-to-depth and normal prior models, and introduces an albedo diffusion model to impose data-driven constraints on the albedo component.
- 212
- 210
- StableNormal: Reducing Diffusion Variance for Stable and Sharp Normal196
StableNormal represents a baby attempt to repurpose diffusion priors for deterministic estimation of surface normal estimation, and evidence that StableNormal retains both the "stability" and "sharpness" for accurate normal estimation is evidence.
- 188
- Conditional Neural Fields174
A new conditional probabilistic graphical model, Conditional Neural Fields (CNF), is proposed, which is the best among approximately 10 machine learning methods for protein secondary structure prediction and also among a few of the best methods for handwriting recognition.
- Unsupervised Template Learning for Fine-Grained Object Recognition166
This work proposes a template model for the purpose of fine-grained recognition, which captures common shape patterns of object parts, as well as the cooccurrence relation of the shape patterns, and achieves significantly outperform the state-of-the-art algorithms.
- 159
- ScratchDet: Training Single-Shot Object Detectors From Scratch141
The ScratchDet achieves the state-of-the-art accuracy on PASCAL VOC 2007, 2012 and MS COCO among all the train-from-scratch detectors and even performs better than several one-stage pretrained methods.
- Detection, Tracking, and Counting Meets Drones in Crowds: A Benchmark140
The Space-Time Neighbor-Aware Network (STNNet) is designed as a strong baseline to solve object detection, tracking and counting jointly in dense crowds and performs favorably against the state-of-the-arts.
- 140
- Social Recommendation with Self-Supervised Metagraph Informax Network132
Self- Supervised Metagraph Informax Network (SMIN) is proposed which investigates the potential of jointly incorporating social- and knowledge-aware relational structures into the user preference representation framework and generalizes the mutual information learning paradigm from vector space to a self-supervised graph-based collaborative filtering.
- 128
- A Scalable Tree-Based Approach for Joint Object and Pose Recognition118
A scalable framework, Object-Pose Tree, which efficiently organizes data into a semantically structured tree, allowing for robust and efficient object category, instance, and pose recognition on challenging everyday objects.
- Graph-Enhanced Multi-Task Learning of Multi-Level Transition Dynamics for Session-based Recommendation113
This paper proposes a multi-task learning framework with Multi-level Transition Dynamics (MTD), which enables the jointly learning of intra- and inter-session item transition dynamics in automatic and hierarchical manner and demonstrates the superiority of MTD as compared to state-of-the-art baselines.
- Toward Robust Material Recognition for Everyday Objects101
This work empirically study material recognition of real-world objects using a rich set of local features using the Kernel Descriptor framework and extends the set of descriptors to include materialmotivated attributes using variances of gradient orientation and magnitude.
- 101
- Compressing Volumetric Radiance Fields to 1 MB100
A robust and adaptive metric for estimating redundancy in grid models and performing voxel pruning by better exploring intermediate outputs of volumetric rendering is presented, and a trainable vector quantization is further proposed to improve the compactness of grid models.
- One-Shot High-Fidelity Talking-Head Synthesis with Deformable Neural Radiance Field99
This paper proposes HiDe-NeRF, which achieves high-fidelity and free-view talking-head synthesis and improves fidelity from two aspects: to enhance identity expressiveness, a generalized appearance module that leverages multi-scale volume features to preserve face shape and details and a lightweight deformation module that explicitly decouples the pose and expression to enable precise expression modeling.
- Generative Adversarial Attributed Network Anomaly Detection98
This paper proposes a generative adversarial attributed network (GAAN) anomaly detection framework and conducts extensive experiments on benchmark datasets and compares with state-of-the-art attributed graph anomaly detection methods.
- Spatiotemporal CNN for Video Object Segmentation98
A unified, end-to-end trainable spatiotemporal CNN model for VOS, which consists of two branches, i.e., the temporal coherence branch and the spatial segmentation branch, which is designed to capture the dynamic appearance and motion cues of video sequences to guide object segmentation.
- Spatial-Temporal Sequential Hypergraph Network for Crime Prediction with Dynamic Multiplex Relation Learning91
To handle spatial-temporal dynamics under the long-range and global context, the proposed ST-SHN framework designs a graph-structured message passing architecture with the integration of the hypergraph learning paradigm and introduces a multi-channel routing mechanism to learn the time-evolving structural dependency across crime types.
- 89
- DiffuEraser: A Diffusion Model for Video Inpainting81
DiffuEraser is introduced, a video inpainting model based on stable diffusion, designed to fill masked regions with greater details and more coherent structures, and outperforms state-of-the-art techniques in both content completeness and temporal consistency while maintaining acceptable efficiency.
- Multi-Behavior Enhanced Recommendation with Cross-Interaction Collaborative Relation Modeling81
A Graph Neural Multi-Behavior Enhanced Recommendation framework which explicitly models the dependencies between different types of user-item interactions under a graph-based message passing architecture and consistently outperforms state-of-the-art methods.
- MIMO: Controllable Character Video Synthesis with Spatial Decomposed Modeling78
Experimental results show that the proposed MIMO method outperforms prior works by a large margin in character animation synthesis and is effective in providing a high degree of controllability, thus enabling brandnew editing tasks (e.g., video character replacement).
- 76
- Multi-Behavior Graph Neural Networks for Recommender System74
A new multi-behavior graph neural network (MBRec), which specially accounts for diverse interaction patterns and the underlying cross-type behavior interdependencies, and develops a graph-structured learning framework to perform expressive modeling of high-order connectivity in behavior-aware user–item interaction graph.
- 74
- 73
- 72
- RGB-D Object Recognition: Features, Algorithms, and a Large Scale Benchmark72
A large-scale, hierarchical multi-view object dataset collected using an RGB-D camera is introduced and it is demonstrated that combining color and depth information substantially improves quality of results.
- 70
- Learning hierarchical sparse features for RGB-(D) object recognition66
HMP builds feature hierarchies layer by layer with an increasing receptive field size to capture abstract representations from raw RGB-D data and indicates that the features learned enable superior object recognition results using linear support vector machines.
- 64
- 56
- EnsemFDet: An Ensemble Approach to Fraud Detection based on Bipartite Graph54
This paper proposes an Ensemble based Fraud DETection (ENSEMFDET) method to scale up fraud detection in bipartite graphs by oversampling the graph and solving the subproblems, the ensemble approach further votes suspicious nodes without sacrificing the prediction accuracy.
- GIC: Gaussian-Informed Continuum for Physical Property Identification and Simulation53
A novel hybrid framework that leverages 3D Gaussian representation to not only capture explicit shapes but also enable the simulated continuum to render object masks as 2D shape surrogates during training is introduced.
- LAM: Large Avatar Model for One-shot Animatable Gaussian Head52
LAM creates an animatable Gaussian head in a single forward pass in seconds, enabling reenactment and rendering without additional networks or post-processing steps, and allows for seamless integration into existing rendering pipelines, ensuring real-time animation and rendering across a wide range of platforms, including mobile phones.
- Deep Dynamic Fusion Network for Traffic Accident Forecasting50
A deep Dynamic Fusion Network framework to explore the central theme of improving the ability of deep neural network on modeling heterogeneous external factors in a fully dynamic manner for traffic accident forecasting and provides a qualitative analysis on prediction results to show the model interpretability.
- SemTalk: Holistic Co-Speech Motion Generation with Frame-Level Semantic Emphasis49
Qualitative and quantitative comparisons demonstrate that the proposed SemTalk method outperforms the state-of-the-art, delivering high-quality co-speech motion with enhanced semantic richness over a stable base motion.
- 47
- Memory-Augmented Non-Local Attention for Video Super-Resolution46
This paper designs a cross-frame non-local attention mechanism that allows video superresolution without frame alignment, leading to being more robust to large motions in the video, and designs a novel memory-augmented attention module to memorize general video details during the superresolution training.
- Adaptive spherical Gaussian kernel in sparse Bayesian learning framework for nonlinear regression43
A novel adaptive spherical Gaussian kernel is utilized for nonlinear regression, and the stagewise optimization algorithm for maximizing Bayesian evidence in sparse Bayesian learning framework is proposed for model selection.
- AniGS: Animatable Gaussian Avatar from a Single Image with Inconsistent Gaussian Reconstruction42
This paper adapts a transformer-based video generation model to generate multi-view canonical pose images and normal maps, pretraining on a large-scale video dataset to improve generalization and proposes a robust method for 3D reconstruction of inconsistent images, enabling real-time rendering during inference.
- 4K-NeRF: High Fidelity Neural Radiance Fields at Ultra High Resolutions40
A novel and effective framework to pursue high fidelity view synthesis on the challenging scenarios of ultra high resolutions, building on the methodology of neural radiance fields, which can significantly boost rendering quality on high-frequency details compared with modern NeRF methods and achieve the state-of-the-art visual quality on 4K ultra-high-resolution scenarios.
- 40
- Image texture classification using a manifold-distance-based evolutionary clustering method38
The experimental results show that in terms of cluster quality and robustness, MEC outperforms the K-means algorithm, a modified K-Means algorithm using the manifold-distance-based dissimilarity measure, and a genetic-algorithm-based clustering technique in partitioning most of the test problems.
- Recursive Finite Newton Algorithm for Support Vector Regression in the Primal37
A recursive finite Newton algorithm (IHLF-SVR-RFN) for training nonlinear support vector regression and the insensitive Huber loss function and the computation of the Newton step are discussed in detail.
- 36
- PanoContext-Former: Panoramic Total Scene Understanding with a Transformer33
A fully 3D method for holistic indoor scene understanding which recovers the objects' shapes, oriented bounding boxes and the 3D room layout simultaneously from a single panorama to maximize the exploration of the rich context information.
- Efficient Pig Counting in Crowds with Keypoints Tracking and Spatial-aware Temporal Response Filtering31
This work presented a real-time automated pig counting system in crowds using only one monocular fisheye camera with an inspection robot, and showed that it achieved performance superior to human.
- RenderIH: A Large-scale Synthetic Dataset for 3D Interacting Hand Pose Estimation28
This work presents a large-scale synthetic dataset –RenderIH– for interacting hands with accurate and diverse pose annotations, and introduces a transformer-based pose estimation network, TransHand, to leverage the correlation between interacting hands and verify the effectiveness of RenderIH in improving results.
- A Modified K-Means Clustering with a Density-Sensitive Distance Metric28
A novel dissimilarity measure, named density-sensitive distance metric, is proposed, which can describe the distribution characteristic of data clustering and is given, which has the ability to identify complex non-convex clusters compared with the original K-Means algorithm.
- SensorSift26
It is found that as long as the public and private attributes are not significantly correlated, it is possible to generate a sifting transformation which reduces private attribute inferences to random guessing while maximally retaining classifier accuracy of public attributes relative to raw data.
- 25
- 24
- 24
- ScratchDet: Exploring to Train Single-Shot Object Detectors from Scratch.24
The ScratchDet achieves the state-of-the-art accuracy on PASCAL VOC 2007, 2012 and MS COCO among all the train-from-scratch detectors and even performs better than several one-stage pretrained methods.
- 23
- 22
- Density-Sensitive Evolutionary Clustering18
The experimental results show that the novel density-sensitive evolutionary clustering algorithm has the ability to identify complex non-convex clusters compared with the K-means algorithm, a genetic algorithm-based clustering, and a modified K-Means algorithm with thedensity-sensitive distance metric.
- Compact RGBD Surface Models Based on Sparse Coding17
This paper decomposes a set of colored point clouds into local surface patches and encodes them based on an overcomplete dictionary and shows that better dictionaries can be learned by extending the K-SVD method with a binary weighting scheme that ignores undefined surface cells.
- OV9D: Open-Vocabulary Category-Level 9D Object Pose and Size Estimation16
Comprehensive quantitative and qualitative experiments demonstrate that the proposed open-vocabulary method, trained on large-scale synthesized data, significantly outperforms the baseline and can effectively generalize to real-world images of unseen categories.
- 16
- 16
- 13
- Supervised Spectral Latent Variable Models12
The resulting Supervised Spectral Latent Variable Model (SSLVM) combines the properties of probabilistic geometric manifold learning (accommodates geometric constraints corresponding to any spectral embedding method including PCA, ISOMAP or Laplacian Eigenmaps), with the additional supervisory information to further constrain it for predictive tasks.
- Synchronized Video-to-Audio Generation via Mel Quantization-Continuum Decomposition11
This paper decomposes the mel-spectrogram into three distinct types of signals, employing quantization or continuity to them, and shows how to balance the representation of mel-spectrograms in terms of completeness and complexity through a new approach called Mel Quantization-Continuum Decomposition (Mel-QCD).
- DiffHand: End-to-End Hand Mesh Reconstruction via Diffusion Models10
DiffHand is developed, the first diffusion-based framework that approaches hand mesh reconstruction as a denoising diffusion process and designs a novel network module called the cross-modality decoder to better model the connectivity of hand vertices.
- Multiple Parameter Selection for LS-SVM Using Smooth Leave-One-Out Error10
SLOO-MPS is proposed for tuning multiple parameters for LS-SVM to overcome the problem of intractable model selection, based on optimizing the smooth leave- one-out error via a gradient descent algorithm and feasible to compute.
- 9
- 9
- 9
- 8
- 8
- 8
- 7
- 7
- 7
- 7
- 7
- 7
- 7
- 6
- 6
- 6
- 6
- 6
- 5
- 5
- 5
- 5
- FaceChain-FACT: Face Adapter with Decoupled Training for Identity-preserved Personalization4
The Face Adapter with deCoupled Training (FACT) framework is proposed, focusing on both model architecture and training strategy, which effectively constrains the effect of face adapters on the facial region, preserving the generative ability of the base model.
- 4
- 4
- 4
- 2
- 2
- 2
- 2
- 2
- 2
- LivingArena: Do LLMs Know What Other LLMs Don't? Peer-Probing as Scalable Evaluation1
Findings show that peer probing can reveal persistent model-specific weaknesses while separately evaluating answering and reliable test construction, and provide a "living"benchmark for model development, red-teaming, and capability-aware multi-agent coordination.
- 1
- 1
- 1
- 1
- Perception-as-Control: Fine-Grained Controllable Image Animation with 3D-Aware Motion Representation1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
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 2026-10-16. 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.