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
- Jelena KovačevićSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works226 from public data
- Actional-Structural Graph Convolutional Networks for Skeleton-Based Action Recognition1,229
The proposed AS-GCN achieves consistently large improvement compared to the state-of-the-art methods and shows promising results for future pose prediction.
- Discrete Signal Processing on Graphs: Sampling Theory692
It is shown that perfect recovery is possible for graph signals bandlimited under the graph Fourier transform and the connection to the sampling theory of finite discrete-time signal processing and previous work on signal recovery on graphs is established.
- V2X-Sim: Multi-Agent Collaborative Perception Dataset and Benchmark for Autonomous Driving419
V2X-Sim seeks to stimulate collaborative perception research for autonomous driving before realistic datasets become widely available and builds an open-source testbed and provides a benchmark for the state-of-the-art collaborative perception algorithms on three tasks, including detection, tracking and segmentation.
- Dynamic Multiscale Graph Neural Networks for 3D Skeleton Based Human Motion Prediction374
Novel dynamic multiscale graph neural networks (DMGNN) to predict 3D skeleton-based human motions and outperforms state-of-the-art methods in both short and long-term predictions on the datasets of Human 3.6M and CMU Mocap.
- Synthesis of a mixed valence state Ce-MOF as an oxidase mimetic for the colorimetric detection of biothiols360
A facile and rapid in situ partial oxidation synthetic strategy for the fabrication of a mixed valence state Ce-MOF (MVCM) which exhibits intrinsic oxidase-like activity is demonstrated and a colorimetric approach for the high-throughput detection of biothiols in serum samples was established.
- Signal Recovery on Graphs: Variation Minimization305
This work formulate graph signal recovery as an optimization problem, for which it provides a general solution through the alternating direction methods of multipliers through which it relates to signal inpainting, matrix completion, robust principal component analysis, and anomaly detection.
- Signal Recovery on Graphs305
This work formulate graph signal recovery as an optimization problem, for which it provides a general solution through the alternating direction methods of multipliers through which it relates to signal inpainting, matrix completion, robust principal component analysis, and anomaly detection.
- Leapfrog Diffusion Model for Stochastic Trajectory Prediction286
The core of the proposed LED is to leverage a trainable leapfrog initializer to directly learn an expressive multi-modal distribution of future trajectories, which skips a large number of denoising steps, significantly accelerating inference speed.
- GroupNet: Multiscale Hypergraph Neural Networks for Trajectory Prediction with Relational Reasoning281
GroupNet, a multiscale hypergraph neural network, is proposed, which is novel in terms of both interaction capturing and representation learning and applies into both CVAE-based prediction system and previous state-of-the-art prediction systems for predicting socially plausible trajectories with relational reasoning.
- 3D Point Cloud Processing and Learning for Autonomous Driving: Impacting Map Creation, Localization, and Perception258
The recent progress in this research area is reviewed and what has been tried and what is needed for practical and safe AVs are summarized.
- Robust Collaborative 3D Object Detection in Presence of Pose Errors247
CoAlign is a novel hybrid collaboration framework that is robust to unknown pose errors and relies on a novel agent-object pose graph modeling to enhance pose consistency among collaborating agents and adopts a multiscale data fusion strategy to aggregate intermediate features at multiple spatial resolutions.
- Symbiotic Graph Neural Networks for 3D Skeleton-Based Human Action Recognition and Motion Prediction240
This work proposes a symbiotic model to handle two tasks jointly, and proposes two scales of graphs to explicitly capture relations among body-joints and body-parts, and shows that the symbiotic graph neural networks achieve better performances on both tasks compared to the state-of-the-art methods.
- Cell clustering for spatial transcriptomics data with graph neural networks229
It is shown that cell clustering for spatial transcripts outperforms other spatial clustering approaches on spatial transcriptomics datasets and can clearly identify all four cell cycle phases from multiplexed error-robust fluorescence in situ hybridization data of cultured cells.
- Remember Intentions: Retrospective-Memory-based Trajectory Prediction218
This work designs a pair of memory banks to explicitly store representative instances in the training set, acting as prefrontal cortex in the neural system, and proposes MemoNet, an instance-based approach that predicts the movement intentions of agents by looking for similar scenarios in theTraining data.
- EqMotion: Equivariant Multi-Agent Motion Prediction with Invariant Interaction Reasoning212
This work proposes Eq-Motion, an efficient equivariant motion prediction model with invariant interaction reasoning module, which achieves state-of-the-art prediction performances on all the four tasks, improving by 24.0/30.1/8.6/9.2%.
- Multivariate Time Series Forecasting With Dynamic Graph Neural ODEs206
A continuous model to forecast multivariate time series forecasting is proposed, designed and solved to complement missing graph topologies and unify both spatial and temporal message passing, allowing deeper graph propagation and fine-grained temporal information aggregation to characterize stable and precise latent spatial-temporal dynamics.
- MotionNet: Joint Perception and Motion Prediction for Autonomous Driving Based on Bird’s Eye View Maps197
Extensive experiments show that the proposed method overall outperforms the state-of-the-arts, including the latest scene-flow- and 3D-object-detection-based methods, indicating the potential value of the proposed methods serving as a backup to the bounding-box-based system, and providing complementary information to the motion planner in autonomous driving.
- Fast Resampling of Three-Dimensional Point Clouds via Graphs182
This work uses a general feature-extraction operator to represent application-dependent features and proposes a general reconstruction error to evaluate the quality of resampling; by minimizing the error, it obtains a general form of optimal resamplings distribution.
- Signal denoising on graphs via graph filtering173
This paper forms graph signal denoising as an optimization problem and derives an exact closed-form solution expressed by an inverse graph filter, as well as an approximate iterative solution express by a standard graph filter.
- FedDisco: Federated Learning with Discrepancy-Aware Collaboration164
This work proposes a novel aggregation method, Federated Learning with Discrepancy-aware Collaboration (FedDisco), whose aggregation weights not only involve both the dataset size and the discrepancy value, but also contribute to a tighter theoretical upper bound of the optimization error.
- 161
- An Extensible Framework for Open Heterogeneous Collaborative Perception159
Heterogeneous ALliance (HEAL), a novel extensible collaborative perception framework that surpasses SOTA methods in performance while reducing the training parameters by 91.5% when integrating 3 new agent types.
- Collaboration Helps Camera Overtake LiDAR in 3D Detection159
The proposed collaborative camera-only 3D detection (CoCa3D) enables agents to share complementary information with each other through communication, and shows a potential that with sufficient collaboration, the camera might overtake LiDAR in some practical scenarios.
- Learning on Attribute-Missing Graphs149
This article makes a shared-latent space assumption on graphs and develops a novel distribution matching-based GNN called structure-attribute transformer (SAT) for attribute-missing graphs that shows better performance than other methods on both link prediction and node attribute completion tasks.
- 140
- SafeAgentBench: A Benchmark for Safe Task Planning of Embodied LLM Agents138
The first comprehensive benchmark for safety-aware task planning of embodied LLM agents in interactive simulation environments, covering both explicit and implicit hazards is presented, covering both explicit and implicit hazards.
- Collaborative Motion Prediction via Neural Motion Message Passing136
This work proposes neural motion message passing (NMMP) to explicitly model the interaction and learn representations for directed interactions between actors, and designs the motion prediction systems for two settings: the pedestrian setting and the joint pedestrian and vehicle setting.
- Signal Recovery on Graphs: Fundamental Limits of Sampling Strategies135
A recovery strategy to compare uniform sampling with experimentally designed sampling and a comprehensive explanation for when and why sampling for semi-supervised learning with graphs works is presented.
- No-Reference Point Cloud Quality Assessment via Domain Adaptation130
A novel no-reference quality assessment metric, the image transferred point cloud quality assessment (IT-PCQA), for 3D point clouds, and introduces a novel con-ditional cross entropy loss in the conditional-discriminative network to penalize the negative samples which hinder the convergence of the quality regression network.
- 129
- Diagnosis algorithms for indirect structural health monitoring of a bridge model via dimensionality reduction112
The results provide evidence for the applicability of indirect structural health monitoring in bridge models and suggest the feasibility of extending this approach to actual structures.
- A Backdoor Attack against 3D Point Cloud Classifiers99
The first backdoor attack (BA) against 3D PC classifiers is proposed, proposing to insert a cluster of points into a PC as a robust backdoor pattern customized for 3D PCs and shows evasiveness against state-of-the-art PC ADs.
- Communication-Efficient Collaborative Perception via Information Filling with Codebook98
Results show that CodeFilling outperforms previous SOTA Where2comm on DAIR-V2X/OPV2VH+ with 1,333/1,206× lower communication volume.
- Interruption-Aware Cooperative Perception for V2X Communication-Aided Autonomous Driving94
This work proposes V2X communication INterruption-aware COoperative Perception (V2X-INCOP), a cooperative perception system robust to communication interruption for V2X communication-aided autonomous driving, which leverages historical cooperation information to recover missing information due to the interruptions and alleviate the impact of the interruption issue.
- Graph Unrolling Networks: Interpretable Neural Networks for Graph Signal Denoising93
The proposed graph unrolling networks expand algorithm unrolling to the graph domain and provide an interpretation of the architecture design from a signal processing perspective and demonstrate that the methods have smaller denoising errors than conventional Denoising algorithms and state-of-the-art graph neural networks.
- 86
- Track-monitoring from the dynamic response of an operational train85
This study examines a more economical approach where track inspection is performed by analyzing vibration data collected from an operational passenger train, and examines how consistently change detection approaches can identify track changes from the data.
- 85
- 80
- Safety at Scale: A Comprehensive Survey of Large Model and Agent Safety76
This survey provides a systematic review of current safety research on large models, covering Vision Foundation Models, Large Language Models, Vision-Language Pre-training models, Vision-Language Models, Diffusion Models, and large-model-powered Agents, and presents a comprehensive taxonomy of safety threats to these models.
- MPED: Quantifying Point Cloud Distortion Based on Multiscale Potential Energy Discrepancy72
By evaluating various neighborhood sizes, the proposed MPED achieves global-local tradeoffs, capturing distortion in a multiscale fashion, and theoretically show that classical Chamfer distance is a special case of the MPED.
- 3D Point Cloud Denoising via Deep Neural Network Based Local Surface Estimation72
Experimental results show that the proposed NPD algorithm can estimate normal vectors of points in noisy point clouds and achieves better denoising performance and produces much smaller variances.
- Among Us: Adversarially Robust Collaborative Perception by Consensus67
The key idea is that collaborative perception should lead to consensus rather than dissensus in results compared to individual perception, which leads to the hypothesize-and-verify framework: perception results with and without collaboration from a random subset of teammates are compared until reaching a consensus.
- Time-variant prediction of flow over an airfoil using deep neural network66
The neural network-based ROM has exhibited attractive potentials on ROM of the unsteady fluid dynamic problem, and the model can potentially serve on investigating flow control or optimization problems in the future.
- Unsteady reduced-order model of flow over cylinders based on convolutional and deconvolutional neural network structure66
A neural network based reduced-order model for predicting the unsteady flow field over single/multiple cylinders achieves fast and accurate prediction on the velocity field, and the CNN-DC decoder model gives more robust and precise performance for all studied problems.
- InfiniCube: Unbounded and Controllable Dynamic 3D Driving Scene Generation with World-Guided Video Models65
InfiniCube is presented, a scalable and controllable method to generate unbounded and dynamic 3D driving scenes with high fidelity and a fast feed-forward approach that employs both voxel and pixel branches to lift videos to dynamic 3D Gaussians with controllable objects.
- Track monitoring from the dynamic response of a passing train: A sparse approach60
A sparse approach to solving an inverse problem where the roughness is constrained to a discrete and limited set of “bumps” and the train system is idealized as a simple damped oscillator that models the train’s vibration in the fundamental mode is examined.
- Online Multi-Agent Forecasting With Interpretable Collaborative Graph Neural Networks59
A novel collaborative prediction unit (CoPU), which aggregates the predictions from multiple collaborative predictors according to a collaborative graph neural network, which has lower average time costs in one online training/testing iteration than most previous methods.
- Generalized Value Iteration Networks:Life Beyond Lattices57
This paper introduces a generalized value iteration network (GVIN), which is an end-to-end neural network planning module that emulates the value iteration algorithm by using a novel graph convolution operator, which enables GVIN to learn and plan on irregular spatial graphs.
- Self-Alignment of Large Language Models via Monopolylogue-based Social Scene Simulation56
This paper presents MATRIX, a novel social scene simulator that emulates realistic scenes around a user's input query, enabling the LLM to take social consequences into account before responding, and fine-tune the LLM with MATRIX-simulated data, ensuring adherence to human values without compromising inference speed.
- Energy-Efficient Computation Offloading in Mobile Edge Computing Systems With Uncertainties54
This paper investigates the energy-efficient computation offloading problem by relaxing common assumptions and considering intrinsic uncertainties in the network, and develops an ϵ-bounded approximation algorithm based on column generation that can efficiently identify a feasible solution that is less than (1+ϵ) of the optimal one.
- TBP-Former: Learning Temporal Bird's-Eye-View Pyramid for Joint Perception and Prediction in Vision-Centric Autonomous Driving52
A temporal bird's-eye-view pyramid transformer (TBP-Former) for vision-centric PnP is proposed, which includes two novel designs that comprehensively extract multi-scale BEV features and predict future BEV states with the support of spatial priors.
- FedFM: Anchor-Based Feature Matching for Data Heterogeneity in Federated Learning52
This work proposes a feature matching loss called contrastive-guiding (CG), which guides each local feature to match with the corresponding anchor while keeping away from non-corresponding anchors, and proposes a FedFM variant, called FedFM-Lite, which enables flexible trade-off between algorithm utility and communication bandwidth cost.
- 51
- A data fusion approach for track monitoring from multiple in-service trains51
Combining data from numerous in-service trains allows for more continuous and more reliable data-driven monitoring than analyzing data from any one train alone; as the number of instrumented trains increases, the proposed fusion approach could facilitate track monitoring of entire rail-networks.
- Auxiliary Tasks Benefit 3D Skeleton-based Human Motion Prediction47
A novel auxiliary-adapted transformer is proposed, which can handle incomplete, corrupted motion data and achieve coordinate recovery via capturing spatial-temporal dependencies through auxiliary tasks, leading to better feature learning.
- 46
- Signal inpainting on graphs via total variation minimization46
A novel recovery algorithm for signals with complex, irregular structure that is commonly represented by graphs is proposed, a generalization of the signal inpainting technique from classical signal processing.
- Aerial Monocular 3D Object Detection45
This work proposes a dual-view detection system named DVDET to achieve aerial monocular object detection in both the 2D image space and the 3D physical space and proposes a novel trainable geo-deformable transformation module that can properly warp information from the drone's perspective to the birds' eye view (BEV).
- Adaptive graph filtering: Multiresolution classification on graphs44
In this work, an adaptive graph filtering approach to semi-supervised classification is presented and the multiresolution property of adaptive graph filters is demonstrated by connecting them to the diffusion wavelets.
- LLM-Powered GUI Agents in Phone Automation: Surveying Progress and Prospects43
This paper systematically reviews LLM-driven phone GUI agents, highlighting their evolution from script-based automation to intelligent, adaptive systems, and proposes a taxonomy covering fundamental agent frameworks, modeling approaches, and essential datasets and benchmarks.
- Toward Collaborative Autonomous Driving: Simulation Platform and End-to-End System43
CoDriving is introduced, a novel end-to-end collaborative driving system that properly integrates V2X communication over the entire autonomous pipeline, promoting driving with shared perceptual information.
- FastMAC: Stochastic Spectral Sampling of Correspondence Graph43
This paper builds a complete 3D registration algorithm termed as FastMAC, that reaches real-time speed while leading to little to none performance drop, and exploits the generalized degree signal on correspondence graph and pursue sampling strategies that preserve highfrequency components of this signal.
- Physical Backdoor Attack can Jeopardize Driving with Vision-Large-Language Models42
BadVLMDriver is proposed, the first backdoor attack against VLMs for autonomous driving that can be launched in practice using physical objects and achieves a 92% attack success rate in inducing a sudden acceleration when coming across a pedestrian holding a red balloon.
- 41
- 39
- PCT: Large-Scale 3d Point Cloud Representations Via Graph Inception Networks with Applications to Autonomous Driving38
This work discretizes the 3D space into voxels and proposes novel graph inception networks to represent 3D points in each voxel, which makes the system avoid discretization errors and work for large-scale scenarios.
- 38
- 37
- Signal recovery on graphs: Random versus experimentally designed sampling36
A new class of smooth graph signals is proposed, called approximately bandlimited, which uses sampling scores, which is similar to the leverage scores used in the matrix approximation for signal recovery on graphs based on random sampling and experimentally designed sampling.
- Grid adaptive reduced-order model of fluid flow based on graph convolutional neural network35
This framework generalizes the graph convolution network to flow field prediction and opens the door to further extending GCN to most existing data-driven architectures of fluid dynamics in the future.
- 34
- V2X-R: Cooperative LiDAR-4D Radar Fusion with Denoising Diffusion for 3D Object Detection32
V2X-R, the first simulated V2X dataset incorporating LiDAR, camera, and 4D radar modalities, is presented and a novel cooperative LiDAR-4D radar fusion pipeline for 3D object detection is proposed and implemented with multiple fusion strategies.
- 31
- 31
- Adaptive Mutual Supervision for Weakly-Supervised Temporal Action Localization30
A novel Adaptive Mutual Supervision framework with two branches, where the base branch detects the most discriminative action regions, while the supplementary branch localizes the less discriminatives through an adaptive sampler, which significantly outperforms state-of-the-art methods.
- Handwritten Chinese Font Generation with Collaborative Stroke Refinement30
The proposed model significantly outperforms the state-of-the-art methods under practical setting, i.e. with only 750 paired training samples, and is added on top of any method for font synthesis.
- Joint-Relation Transformer for Multi-Person Motion Prediction29
This paper proposes the Joint-Relation Transformer, which utilizes relation information to enhance interaction modeling and improve future motion prediction, and designs a novel joint-relation fusion layer with relation-aware attention to update both features.
- PointMotionNet: Point-Wise Motion Learning for Large-Scale LiDAR Point Clouds Sequences29
This work proposes a point-based spatiotemporal pyramid architecture, called PointMotionNet, to learn motion information from a sequence of large-scale 3D LiDAR point clouds, and designs an end-to-end system where Point motionNet is the core module that learns motion information.
- Distributed algorithm for graph signal inpainting29
Experiments on online blog classification and temperature prediction suggest that the convergence speed of the proposed distributed algorithm is competitive with that of the centralized algorithm, especially when a graph tends to be regular.
- The eigen-structures of real (skew) circulant matrices with some applications28
A DCT-DST version of circulant and skew-circulant splitting (CSCS) iteration for real positive definite Toeplitz systems is developed, which is more efficient and saves a half storage.
- Pruned Graph Scattering Transforms27
This work addresses some limitations of GSTs by introducing a novel so-termed pruned (p)GST approach, which is guided by a graph-spectrum-inspired criterion, and retains informative scattering features on-the-fly while bypassing the exponential complexity associated with GSTs.
- Pragmatic Communication in Multi-Agent Collaborative Perception26
This paper proposes PragComm, a multi-agent collaborative perception system with two key components: i) single-agent detection and tracking and ii) pragmatic collaboration that promotes pragmatic communication and adapts to a wide range of communication conditions.
- Hypergraph Transformer for Semi-Supervised Classification26
This work proposes a novel hypergraph learning framework, HyperGraph Transformer, which uses a Transformer-based neural network architecture to effectively consider global correlations among all nodes and hyperedges and achieves comprehensive hypergraph representation learning by effectively incorporating global interactions while preserving local connectivity patterns.
- Emerging Safety Attack and Defense in Federated Instruction Tuning of Large Language Models26
This paper reveals the vulnerability of safety alignment in FedIT by proposing a simple, stealthy, yet effective safety attack method and proposes a post-hoc defense method, which could rely on a fully automated pipeline: generation of defense data and further fine-tuning of the LLM.
- Discriminative Radial Domain Adaptation26
This paper proposes Discriminative Radial Domain Adaptation (DRDA), a method which bridges source and target domains via a shared radial structure to reduce domain shift via structure matching and enhance feature transferability and discriminability simultaneously.
- ML-Agent: Reinforcing LLM Agents for Autonomous Machine Learning Engineering25
This work proposes a novel agentic ML training framework with three key components: an exploration-enriched fine-tuning, which enables LLM agents to generate diverse actions for enhanced RL exploration; a step-wise RL, which enables training on a single action step, accelerating experience collection and improving training efficiency.
- 25
- CaT: Weakly Supervised Object Detection with Category Transfer24
A novel category transfer framework for weakly supervised object detection that achieves 63.5% mAP and 80.3% CorLoc with 5 overlapping categories between two datasets, which outperforms the state-of-the-art methods.
- 24
- 23
- Neural Message Passing for Visual Relationship Detection21
This work explicitly model objects and interactions by an interaction graph and then proposes a message-passing-style algorithm to propagate the contextual information and calls the proposed method neural message passing (NMP).
- Invariant Teacher and Equivariant Student for Unsupervised 3D Human Pose Estimation21
A novel method based on teacher-student learning framework for 3D human pose estimation without any 3D annotation or side information is proposed and reduces the 3D joint prediction error compared to state-of-the-art unsupervised methods and also outperforms many weakly- supervised methods that use side information on Human3.6M.
- Dynamic responses, GPS positions and environmental conditions of two light rail vehicles in Pittsburgh21
The DR-Train dataset provides dynamic responses of in-service trains via vibration data collected by accelerometers, which enables a low-cost way of monitoring rail tracks more frequently and will result in more reliable and economical ways to monitor rail infrastructure.
- Collaborative Uncertainty Benefits Multi-Agent Multi-Modal Trajectory Forecasting20
This work proposes a novel concept, collaborative uncertainty (CU), which models the uncertainty resulting from interaction modules, and builds a general CU-aware regression framework with an original permutation-equivariant uncertainty estimator to do both tasks of regression and uncertainty estimation.
- Robust Collaborative Perception without External Localization and Clock Devices19
This work proposes a robust collaborative perception system that operates independently of external localization and clock devices called FreeAlign, which constructs a salient object graph for each agent based on its detected boxes and uses a graph neural network to identify common subgraphs between agents, leading to accurate relative pose and time.
- 19
- 19
- One is Plenty: A Polymorphic Feature Interpreter for Immutable Heterogeneous Collaborative Perception18
PolyInter provides an extension point where new agents integrate by overriding only their specific prompts, which are learnable parameters that guide interpretation, while reusing PolyInter’s remaining parameters to achieve both extensibility in immutable heterogeneous scenarios and low-loss feature interpretation.
- V2X-PC: Vehicle-to-everything Collaborative Perception via Point Cluster18
This work introduces a brand new message unit, namely point cluster, designed to represent the scene sparsely with a combination of low-level structure information and high-level semantic information, and proposes a novel framework V2X-PC for collaborative perception.
- Wireless 3D Point Cloud Delivery Using Deep Graph Neural Networks18
The proposed scheme introduces an end-to-end deep learning framework based on graph neural network (GNN) to reconstruct high-quality point clouds from its distorted observation under wireless fading channels to simultaneously realize better 3D reconstruction quality and lower communication overheads.
- 18
- 18
- Detecting Backdoor Attacks against Point Cloud Classifiers17
A reverse-engineering defense is proposed that infers whether a PC classifier is backdoor attacked, without access to its training set or to any clean classifiers for reference, and is demonstrated on the benchmark ModeNet40 dataset for PCs.
- 17
- Multiresolution classification with semi-supervised learning for indirect bridge structural health monitoring16
This work presents a multiresolution classification framework with semi-supervised learning for the indirect structural health monitoring of bridges and shows that the proposed algorithm performs significantly better than supervised multiresolved classification.
- BrowseMaster: Towards Scalable Web Browsing via Tool-Augmented Programmatic Agent Pair15
The proposed BrowseMaster is a scalable framework built around a programmatically augmented planner-executor agent pair, which demonstrates its strong capability in complex, reasoning-heavy information-seeking tasks at scale.
- 15
- A statistical perspective of sampling scores for linear regression15
This paper proposes a simple yet effective estimator, which is easy for theoretical analysis and is useful in multitask linear regression, and derives the exact mean square error of the proposed estimator for any given sampling scores.
- 14
- Compatible Transformer for Irregularly Sampled Multivariate Time Series14
The proposed CoFormer is a transformer-based encoder to achieve comprehensive temporal-interaction feature learning for each individual sample in irregular multivariate time series, and can analyze irregularly sampled multivariate time series for many downstream tasks, including classification and prediction.
- 13
- Malicious Agent Detection for Robust Multi-Agent Collaborative Perception13
This paper proposes Malicious Agent Detection (MADE), a reactive defense specific to MAC perception that can be deployed by each agent to accurately detect and then remove any potential malicious agent in its local collaboration network.
- Fake It Till Make It: Federated Learning with Consensus-Oriented Generation13
This paper proposes federated learning with consensus-oriented generation (FedCOG), which can be a plug-and-play module to further improve the performance of most existing FL methods and is naturally compatible with standard FL protocols such as Secure Aggregation since it makes no modification in communication process.
- A 3D Mesh-Based Lifting-and-Projection Network for Human Pose Transfer13
Experiments empirically demonstrate that the proposed lifting-and-projection framework is effective and outperforms the existing image-to-image-based and mesh-based methods on human pose transfer task in both self-transfer and cross-transfer settings.
- Damage-Sensitive and Domain-Invariant Feature Extraction for Vehicle-Vibration-Based Bridge Health Monitoring13
A physics-guided signal processing approach to extract a damage-sensitive and domain-invariant DS & DI feature from acceleration response data of a vehicle traveling over a bridge to assess bridge health is introduced.
- 12
- Learning Hypergraphs From Signals With Dual Smoothness Prior12
This paper proposes a hypergraph structure learning framework HGSL with a novel dual smoothness prior that reveals a mapping between the observed node signals and the hyper graph structure, whereby each hyperedge corresponds to a subgraph with both node signal smoothness and edge signal Smoothness in the learnable graph structure.
- CCST: Cell clustering for spatial transcriptomics data with graph neural network12
CCST is a general framework for dealing with various kinds of spatially resolved transcriptomics that outperforms other spatial cluster approaches on spatial transcriptomics datasets, and can clearly identify all four cell cycle phases from MERFISH data of cultured cells.
- 12
- 12
- MobileA3gent: Training Mobile GUI Agents Using Decentralized Self-Sourced Data from Diverse Users11
The proposed MobileA3gent is a collaborative framework that trains mobile GUI Agents using decentralized self-sourced data from diverse users and achieves superior performance over traditional approaches at only 1% of the cost, highlighting its potential for real-world applications.
- Hypergraph-Mlp: Learning on Hypergraphs Without Message Passing11
This work introduces Hypergraph-MLP, a novel learning framework for hypergraph-structured data, where the learning model is a straightforward multilayer perceptron (MLP) supervised by a loss function based on a notion of signal smoothness on hypergraphs.
- 11
- 10
- 10
- 10
- FedRSU: Federated Learning for Scene Flow Estimation on Roadside Units9
FedRSU is introduced, an innovative federated learning framework for self-supervised scene flow estimation that presents a recurrent self-supervision training paradigm, where for each RSU, the scene flow prediction of points at every timestamp can be supervised by its subsequent future multi-modality observation.
- 9
- 9
- 9
- 9
- Detecting Localized Categorical Attributes on Graphs9
This paper proposes two statistics: Graph wavelet statistic and graph scan statistic, both of which are provably effective in detecting localized attributes in categorical attributes such as attributes in online social networks, ratings in recommender systems, and viruses in cyber-physical systems.
- 9
- 8
- 8
- Signal Localization, Decomposition and Dictionary Learning on Graphs.8
This work considers three critical problems in graph signal processing: localization, decomposition and dictionary learning, and proposes a specific graph signal model, an optimization problem and a computationally efficient solver for each.
- 8
- Self-Localized Collaborative Perception7
A novel self-localized collaborative perception system, which achieves more holistic and robust collaboration without using an external localization system, and a novel spatial alignment module, which provides the relative poses between agents by effectively matching co-visible objects across agents.
- Hierarchical Spherical CNNs With Lifting-Based Adaptive Wavelets for Pooling and Unpooling6
This work proposes a novel framework of HS-CNNs with lifting structures to learn adaptive spherical wavelets for pooling and unpooling, named LiftHS-CNNs, and proposes an unpooling operation invertible to the lifting-based pooling for restoring the up-scaled representations, which can well preserve spectral characteristics of the original signal.
- Enhancing Data Quality in Federated Fine-Tuning of Foundation Models6
A data quality control pipeline for federated fine-tuning of foundation models is proposed, which computes scores reflecting the quality of training data and determines a global threshold for a unified standard, aiming for improved global performance.
- 6
- 6
- 6
- Collaborative Joint Perception and Prediction for Autonomous Driving6
The experimental results demonstrate that CoPnP outperforms existing state-of-the-art collaboration methods, achieving a significant performance-communication trade-off and yielding up to 11.51%/10.34% Intersection over union and 12.31%/10.96% video panoptic quality gains over single-agent PnP on the OPV2V/V2XSet datasets.
- 6
- 5
- 5
- Federated Learning Empowered by Generative Content5
A novel FL framework termed FedGC is proposed, designed to mitigate data heterogeneity issues by diversifying private data with generative content by generating data based on the guidance of prompts and real data simultaneously.
- 5
- Semi-Supervised 3D Hand-Object Pose Estimation Via Pose Dictionary Learning5
A semi-supervised 3D hand-object pose estimation method with two key techniques: pose dictionary learning and an object-oriented coordinate system that can distinguish infeasible poses by reconstruction error and make 3D estimations equivariant to the camera perspective.
- 5
- 5
- 5
- Detecting Structure-correlated Attributes on Graphs.5
This paper forms a statistical hypothesis testing to decide if the given attribute activates a community in a graph with interfering by the Bernoulli noise and proposes two statistics: graph wavelet statistic and graph scan statistic that are effective and efficient to detect activations.
- 4
- 4
- Learn What You Need in Personalized Federated Learning4
A novel algorithm-unrolling-based personalized federated learning framework, enabling each client to adaptively select which part of its local model parameters should participate in collaborative training via algorithm unrolling methods, which significantly outperforms previous personalized federated learning methods.
- 4
- Ontogenetic color change in the tail of blue‐tailed skinks (Plestodion elegans)4
While this study does not provide direct evidence to explain the potential factors that drive dichromatism between the sexes of lizards, it could serve as a reference for future studies exploring possible mechanisms of ontogenetic color change in reptiles.
- Spatio-Temporal Graph Complementary Scattering Networks4
A novel complementary mechanism to organically combine the spatio-temporal graph scattering transform and neural networks is proposed, resulting in the proposed ST-GCSN, which outperforms both ST-GCN andST-GST on hand pose action recognition.
- Task-Level Aware Scheduling of Energy-Constrained Applications on Heterogeneous Multi-Core System4
A novel task energy consumption pre-allocation strategy that is conducive to minimizing the scheduling time and a novel task schedule algorithm based on it is developed and can reach better performance compared with the existing algorithms.
- Dual Graph Embedding for Object-Tag Link Prediction on the Knowledge Graph4
A novel approach named Dual Graph Embedding (DGE) is proposed that models both the first-order and high-order proximities in the KG via an auto-encoding architecture to facilitate better object-tag relation inference and outperforms the state-of-the-art methods.
- 4
- 4
- Fast Temporal Path Localization on Graphs via Multiscale Viterbi Decoding4
This work considers a problem of localizing a temporal path signal that evolves over time on a graph in a series of consecutive time stamps, and proposes a path-localization algorithm with significantly reduced computational complexity.
- 3
- 3
- 3
- EmboMatrix: A Scalable Training-Ground for Embodied Decision-Making3
This work presents EmboMatrix, the first training ground of its kind, providing massive and diverse tasks with efficient simulation and precise rewards, and cultivates EmboBrain, an LLM whose embodied decision-making abilities emerge from extensive embodied interactions.
- 3
- Data Quality Control in Federated Instruction-tuning of Large Language Models3
FedDQC, a novel federated instruction tuning framework with dynamic data quality control, designs a quality-aware hierarchical FL training framework, where the LLM is progressively fine-tuned from high- to low-IRA data in a collaborative manner, enabling dynamic adjustments throughout the training process.
- 3
- Joint 3D Human Shape Recovery and Pose Estimation from a Single Image with Bilayer Graph3
A dual-scale graph approach that uses a coarse graph, derived from a dense graph, and the dense graph to estimate the human’s 3D pose, and introduces graph fusion blocks to exchange information between graphs with different scales.
- 3
- 3
- Collaborative Adversarial Learning for Relational Learning on Multiple Bipartite Graphs3
This paper proposes Collaborative Adversarial Learning (CAL), a model that explicitly models the joint distribution of the shared entities across multiple bipartite graphs and consists of distribution-level and feature-level alignments for knowledge from multiple bipartsite graphs that outperforms the existing methods.
- 3
- Unveiling the Impact of Data and Model Scaling on High-Level Control for Humanoid Robots2
This work introduces Humanoid-Union, a large-scale dataset generated through an autonomous pipeline, comprising over 260 hours of diverse, high-quality humanoid robot motion data with semantic annotations derived from human motion videos, and proposes SCHUR, a scalable learning framework designed to explore the impact of large-scale data on high-level control in humanoid robots.
- Agentic Time Machine as an Infrastructure for Future-Event Forecasting2
This work introduces Agentic Time Machine, an infrastructure that approximately reconstructs the web state at any chosen past time by filtering post-cutoff content and proposes a planner-solver-aggregator multi-agent framework that breaks each question into diverse analytical angles, gathers evidence in parallel, and combines the results into a single forecast.
- 2
- RoCo-Sim: Enhancing Roadside Collaborative Perception through Foreground Simulation2
RoCo-Sim significantly improves roadside 3D object detection, outperforming SOTA methods by 83.74% on Rcooper-Intersection and 83.12% on TUMTraf-V2X for AP70.
- 2
- 2
- 2
- Unrolled Graph Learning for Multi-Agent Collaboration2
This work proposes a distributed multi-agent learning model inspired by human collaboration, in which the agents can autonomously detect suitable collaborators and refer to collaborators' model for better performance, and designs a graph learning network by unrolling, which can learn underlying similar features among potential collaborators.
- 2
- 2
- 2
- 2
- 2
- VeloTrace Reconciles Divergent Velocity and Trajectory in Single-cell Transcriptomics with Deep Neural ODE1
VeloTrace uniquely reconciles RNA velocity and trajectory inference, creating a velocity field where each cell can infer past and future transitions from its current state, and lays the foundation for more accurate modeling of gene regulation and cell fate decisions in complex biological systems.
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- –
- –
- –
- –
- –
- –
- –
- Graph Transductive Defense: a Two-Stage Defense for Graph Membership Inference Attacks–
This paper proposes an effective two-stage defense, Graph Transductive Defense (GTD), tailored to graph transductive learning characteristics, and demonstrates its potential for seamless integration into various classification models with minimal overhead.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Fast algorithm for neural network reconstruction–
An efficient and accurate way of predicting the connectivity of neural networks in the brain represented by simulated calcium fluorescence data is proposed based on matrix completion and a local thresholding technique.
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.