Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks90 from public data
- 1,435
- DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusion1,203
DenseFusion is a generic framework for estimating 6D pose of a set of known objects from RGB-D images that processes the two data sources individually and uses a novel dense fusion network to extract pixel-wise dense feature embedding, from which the pose is estimated.
- AnyGrasp: Robust and Efficient Grasp Perception in Spatial and Temporal Domains542
This article proposes AnyGrasp for grasp perception to enable robots to grasp as robustly as humans using a parallel gripper and develops a dense supervision strategy with real perception and analytic labels in the spatial–temporal domain.
- Virtual to Real Reinforcement Learning for Autonomous Driving356
A novel realistic translation network is proposed to make model trained in virtual environment be workable in real world, and is believed to be the first successful case of driving policy trained by reinforcement learning that can adapt to real world driving data.
- RH20T: A Comprehensive Robotic Dataset for Learning Diverse Skills in One-Shot288
This paper aims to unlock the potential for an agent to generalize to hundreds of real-world skills with multi-modal perception with a dataset comprising over 110,000 contact-rich robot manipulation sequences across diverse skills, contexts, robots, and camera viewpoints, all collected in the real world.
- Towards a new generation of artificial intelligence in China285
This Perspective briefly introduces the New Generation Artificial Intelligence Development Plan of China (2015–2030) from the point of view of the authors, a group of AI experts from academia and industry who have been involved in various stages of the plan.
- EmbodiedScan: A Holistic Multi-Modal 3D Perception Suite Towards Embodied AI206
This work introduces EmbodiedScan, a multi-modal, ego-centric 3D perception dataset and benchmark for holistic 3D scene understanding, and a baseline framework named Embodied Perceptron, capable of processing an arbitrary number of multi-modal inputs and demonstrating remarkable 3D perception capabilities.
- Symmetry and Group in Attribute-Object Compositions163
This paper proposes a previously ignored principle of attribute-object transformation: Symmetry and builds a transformation framework inspired by group theory, i.e. SymNet, which can be utilized for the Compositional Zero-Shot Learning task and outperforms the state-of-the-art on widely-used benchmarks.
- Detailed 2D-3D Joint Representation for Human-Object Interaction158
A detailed 2D-3D joint representation learning method for human-Object Interaction detection and a new benchmark named Ambiguous-HOI consisting of hard ambiguous images are proposed to better evaluate the 2D ambiguity processing capacity of models.
- Asynchronous Interaction Aggregation for Action Detection139
The Asynchronous Interaction Aggregation network (AIA) is proposed that leverages different interactions to boost action detection and can gain notable accuracy from the integrative interactions and is easy to train end-to-end.
- ArtiBoost: Boosting Articulated 3D Hand-Object Pose Estimation via Online Exploration and Synthesis121
This work proposes ArtiBoost, a lightweight online data enhancement method that performs data exploration and synthesis within a learning pipeline, and those synthetic data are blended into real-world source data for training.
- FoAR: Force-Aware Reactive Policy for Contact-Rich Robotic Manipulation83
FoAR is proposed, a force-aware reactive policy that combines high-frequency force/torque sensing with visual inputs to enhance the performance in contact-rich manipulation.
- Upcycling Models Under Domain and Category Shift75
This paper designs a novel, adaptive one-vs-all global clustering algorithm to achieve the distinction across different target classes and introduces a local k-NN clustering strategy to alleviate negative transfer.
- 66
- Real-time contrast preserving decolorization63
Decolorization methods are typically orders of magnitude slower than standard procedures such as matlab rgb2gray function, which largely hinders their practical use.
- Deep RNN Framework for Visual Sequential Applications56
This work proposes a new recurrent neural framework that can be stacked deep effectively and provides empirical evidence to show that the deep RNN framework is easy to optimize and can gain accuracy from the increased depth on several visual sequence problems.
- AirExo-2: Scaling up Generalizable Robotic Imitation Learning with Low-Cost Exoskeletons46
AirExo-2 is presented, a low-cost exoskeleton system for large-scale in-the-wild data collection, along with several adaptors that transform collected data into pseudo-robot demonstrations suitable for policy learning, and RISE-2, a generalizable imitation learning policy that fuses 3D spatial and 2D semantic perception for robust manipulations.
- 46
- 44
- Skeleton-Based Mutually Assisted Interacted Object Localization and Human Action Recognition42
A joint learning framework for mutually assisted “interacted object localization” and “human action recognition” based on skeleton data is proposed and the temporal consistency of interacted object is explored as constraint to better localize the interacted object with the absence of ground-truth labels.
- 42
- 41
- Approximated Bilinear Modules for Temporal Modeling33
It is shown how two-layer subnets in CNNs can be converted to temporal bilinear modules by adding an auxiliary-branch, and snippet sampling and shifting inference are introduced to boost sparse-frame video classification performance.
- Complex sequential understanding through the awareness of spatial and temporal concepts28
A new modelling strategy—‘semi-coupled structure’ (SCS)—which consists of deep neural networks that decouple the complex spatial and temporal concepts during learning is introduced, demonstrating that SCS has the capacity to improve the performance of long short-term memory (LSTM)-like models on large-scale sequential tasks.
- 26
- Template-Instance Loss for Offline Handwritten Chinese Character Recognition26
A character template is designed to deal with the intrinsic similarities among Chinese characters and the instance loss can reduce category variance according to classification difficulty, giving a large penalty to the outlier instance of handwritten Chinese character.
- AXNet: approximate computing using an end-to-end trainable neural network25
This paper proposes a novel neural network structure—AXNet—to fuse two NNs to a holistic end-to-end trainable NN that can tremendously improve the invocation and reduce the approximation error.
- ImplicitRDP: An End-to-End Visual-Force Diffusion Policy With Structural Slow-Fast Learning24
This work proposes ImplicitRDP, a unified end-to-end visual-force diffusion policy that integrates visual planning and reactive force control within a single network, and introduces Structural Slow-Fast Learning, a mechanism utilizing causal attention to simultaneously process asynchronous visual and force tokens.
- TacIPC: Intersection- and Inversion-Free FEM-Based Elastomer Simulation for Optical Tactile Sensors24
This work adopts the finite element method (FEM) and addresses the challenges in penetration and mesh distortion with incremental potential contact (IPC) method and presents a simulator named TacIPC, which can ensure numerically stable simulations while accommodating direct rendering and friction modeling.
- 24
- 23
- 21
- Fast Abnormal Event Detection21
An efficient sparse combination learning framework with both batch and online solvers that reaches high detection rates on benchmark datasets at a speed of 1000–1200 frames per second on average when computing on an ordinary single core desktop PC using MATLAB.
- $L_0$ Regularized Stationary-Time Estimation for Crowd Analysis21
A robust algorithm for estimating how long a foreground pixel becomes stationary is proposed, formulated as an L0 optimization problem, which distinguishes among different foreground objects, which are close or overlapped in the spatio-temporal space by using a locally shared foreground codebook.
- Low-Rank Similarity Mining for Multimodal Dataset Distillation20
This work proposes Low-Rank Similarity Mining (LoRS) for multimodal dataset distillation, that concurrently distills a ground truth similarity matrix with image-text pairs, and leverages low-rank factorization for efficiency and scalability.
- 17
- Differentiable Cloth Parameter Identification and State Estimation in Manipulation17
DiffCP is introduced, a novel differentiable pipeline that leverages the Anisotropic Elasto-Plastic (A-EP) constitutive model, tailored for differentiable computation and robotic tasks, and adopts a “real-to-sim-to-real” methodology.
- Highlighting Object Category Immunity for the Generalization of Human-Object Interaction Detection17
A novel metric is proposed, mPD (mean Performance Degradation), as a complementary of mAP to evaluate the performance gap among compositions of different objects and the same verb, and reveals that previous methods usually generalize poorly.
- ArtGS: 3D Gaussian Splatting for Interactive Visual-Physical Modeling and Manipulation of Articulated Objects16
This work introduces ArtGS, a novel framework that extends 3D Gaussian Splatting (3DGS) by integrating visual-physical modeling for articulated object understanding and interaction, and establishes a new framework for efficient, scalable, and generalizable articulated object modeling and manipulation.
- DexTOG: Learning Task-Oriented Dexterous Grasp With Language Condition16
A novel language-guided diffusion-based learning framework, DexTOG, aimed at advancing the field of task-oriented grasping with dexterous hands is introduced, which includes a diffusion-based grasp pose generation model, DexDiffu, and a data engine to support the DexDiffu.
- UniAff: A Unified Representation of Affordances for Tool Usage and Articulation with Vision-Language Models15
This work proposes a comprehensive paradigm, termed UniAff, that integrates 3D object-centric manipulation and task understanding in a unified formulation and constructed a dataset labeled with manipulation-related key attributes, comprising 900 articulated objects and 600 tools from 12 categories.
- 15
- Human Action Adverb Recognition: ADHA Dataset and a Three-Stream Hybrid Model15
An in-depth analysis on the implementation of current effective models in action recognition and image captioning on adverb recognition is committed, and the results reveal that such methods are unsatisfactory.
- GLC++: Source-Free Universal Domain Adaptation Through Global-Local Clustering and Contrastive Affinity Learning14
This paper proposes a novel Global and Local Clustering (GLC) technique, which comprises an adaptive one-vs-all global clustering algorithm to discern between target classes, complemented by a local k-NN clustering strategy to mitigate negative transfer.
- TieBot: Learning to Knot a Tie from Visual Demonstration through a Real-to-Sim-to-Real Approach14
This work presents TieBot, a Real-to-Sim-to-Real learning from visual demonstration system for the robots to learn to knot a tie, and introduces the Hierarchical Feature Matching approach to estimate a sequence of tie's meshes from the demonstration video.
- RFTrans: Leveraging Refractive Flow of Transparent Objects for Surface Normal Estimation and Manipulation14
RFTrans, an RGB-D-based method for surface normal estimation and manipulation of transparent objects and a real-world robot grasping task witnesses an 83% success rate, proving that refractive flow can help enable direct sim-to-real transfer.
- PGT: A Progressive Method for Training Models on Long Videos13
This work proposes to treat videos as serial fragments satisfying Markov property, and train it as a whole by progressively propagating information through the temporal dimension in multiple steps, able to train long videos end-to-end with limited resources and ensures the effective transmission of information.
- Learning Single/Multi-Attribute of Object With Symmetry and Group13
A Relative Moving Distance (RMD) based method to utilize the attribute change instead of the attribute pattern itself to classify attributes and is suitable for complex compositions of multiple attributes and objects when incorporating attribute correlations.
- CPPF++: Uncertainty-Aware Sim2Real Object Pose Estimation by Vote Aggregation12
This paper introduces a novel method, CPPF++, designed for sim-to-real category-level pose estimation, which significantly surpasses previous sim-to-real approaches and achieves comparable or superior performance on novel datasets.
- Intersection-Free Robot Manipulation With Soft-Rigid Coupled Incremental Potential Contact12
A novel simulation platform designed for robotic manipulation tasks concerning soft objects, ZeMa prioritizes physical accuracy and integrates the incremental potential contact method, offering unified dynamics simulation for both soft and rigid objects.
- Distill Gold from Massive Ores: Bi-level Data Pruning Towards Efficient Dataset Distillation12
This work studies the data efficiency and selection for the dataset distillation task, and proposes to use the empirical loss value as a static data pruning criterion to compensate for the variation of the data value in training.
- 12
- Unsupervised Representation for Semantic Segmentation by Implicit Cycle-Attention Contrastive Learning10
The cycle-attention contrastive learning (CACL) makes use of semantic continuity of video frames, adopting unsupervised cycle-consistent attention mechanism to implicitly conduct contrastivelearning with difficult, global-local-balanced positive pixel pairs.
- 9
- 9
- TrajBooster: Boosting Humanoid Whole-Body Manipulation via Trajectory-Centric Learning8
TrajBooster is presented, a cross-embodiment framework that leverages abundant wheeled-humanoid data to boost bipedal VLA performance, reducing reliance on costly same-embodiment data while enhancing action space understanding and zero-shot skill transfer capabilities.
- DiPGrasp: Parallel Local Searching for Efficient Differentiable Grasp Planning8
This work presents DiPGrasp, a grasp planner that takes a force-closure geometric surface matching grasp quality metric, and adopts a gradient-based optimization scheme on the metric, which also considers parallel sampling and collision handling.
- 8
- 8
- Dynamic Context Removal: A General Training Strategy for Robust Models on Video Action Predictive Tasks8
A simple but effective training strategy, Dynamic Context Removal (DCR), which dynamically schedules the visibility of context in different training stages, and explores how to train robust models that give consistent predictions at different levels of observable context.
- 8
- FAVOR: Full-Body AR-Driven Virtual Object Rearrangement Guided by Instruction Text7
This study introduces FAVOR: a novel dataset for Full-body AR-driven Virtual Object Rearrangement that uniquely employs motion capture systems and AR eyeglasses and presents a pipeline FAVORITE for producing digital human rearrangement motion sequences guided by instructions.
- Learning Dexterous Manipulation with Quantized Hand State6
DQ-RISE is proposed, which quantizes hand states to simplify hand motion prediction while preserving essential patterns, and applies a continuous relaxation that allows arm actions to diffuse jointly with these compact hand states, enabling the policy to learn arm-hand coordination from data while preventing hand actions from overwhelming the action space.
- FA-RDP: A Frequency-Adaptive Reactive Diffusion Policy for Contact-Rich Manipulation4
FA-RDP is presented, a frequency-adaptive reactive diffusion policy that reparameterizes the diffusion network to predict actions on the robot action manifold while retaining DDPM-based residual supervision, and Manifold Consistency Distillation is introduced.
- 4
- 4
- 4
- 4
- 3
- M 3 -VOS: Multi-Phase, Multi-Transition, and Multi-Scenery Video Object Segmentation3
The inherent changes in disorder suggest that the predictive performance of the forward entropy-increasing process can be improved through a reverse entropy-reducing process, which leads to ReVOS, a new plug-and-play model that improves its performance by reversal refinement.
- Dexterous Manipulation Based on Prior Dexterous Grasp Pose Knowledge3
This work introduces a novel reinforcement learning approach that leverages prior dexterous grasp pose knowledge to enhance both efficiency and accuracy in dexterous manipulation.
- DISCO: Embodied Navigation and Interaction via Differentiable Scene Semantics and Dual-Level Control3
- 3
- 3
- 2
- Clutter suppression method in GPR using particle clustering2
A novel clutter suppression method in Ground Penetrating Radar (GPR) that outperforms existing approaches by suppressing time segments of hill represented by their corresponding particle in B-scan.
- 1
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Statistical approximation based fine frequency synchronization for OFDM systems–
A novel approach for fine frequency synchronization of OFDM synchronization systems in multi-path channels with very robust estimation of frequency offset, and a coarse timing offset estimation.
- –
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.