Yong Liu

49

Papers

580

Total Citations

Papers (49)

ToolACE: Winning the Points of LLM Function Calling

Paint3D: Paint Anything 3D with Lighting-Less Texture Diffusion Models

DriveArena: A Closed-loop Generative Simulation Platform for Autonomous Driving

Sundial: A Family of Highly Capable Time Series Foundation Models

REEF: Representation Encoding Fingerprints for Large Language Models

Universal Segmentation at Arbitrary Granularity with Language Instruction

LLaVA-KD: A Framework of Distilling Multimodal Large Language Models

Densely Connected Parameter-Efficient Tuning for Referring Image Segmentation

MMAD: A Comprehensive Benchmark for Multimodal Large Language Models in Industrial Anomaly Detection

Beyond Prototypes: Semantic Anchor Regularization for Better Representation Learning

Towards a Theoretical Understanding of Synthetic Data in LLM Post-Training: A Reverse-Bottleneck Perspective

InstructSeg: Unifying Instructed Visual Segmentation with Multi-modal Large Language Models

Flash-VStream: Efficient Real-Time Understanding for Long Video Streams

OLinear: A Linear Model for Time Series Forecasting in Orthogonally Transformed Domain

NeurIPS 2025arXiv

A Multimodal, Multi-Task Adapting Framework for Video Action Recognition

TIMotion: Temporal and Interactive Framework for Efficient Human-Human Motion Generation

SAM2-LOVE: Segment Anything Model 2 in Language-aided Audio-Visual Scenes

Understanding Fairness Surrogate Functions in Algorithmic Fairness

LORS: Low-rank Residual Structure for Parameter-Efficient Network Stacking

CoHD: A Counting-Aware Hierarchical Decoding Framework for Generalized Referring Expression Segmentation

HyperSeg: Hybrid Segmentation Assistant with Fine-grained Visual Perceiver

OvSW: Overcoming Silent Weights for Accurate Binary Neural Networks

Understanding Model Ensemble in Transferable Adversarial Attack

Decentralized Federated Learning with Model Caching on Mobile Agents

High-Dimensional Analysis for Generalized Nonlinear Regression: From Asymptotics to Algorithm

Action Detail Matters: Refining Video Recognition with Local Action Queries

Stepping Out of Similar Semantic Space for Open-Vocabulary Segmentation

PatchScaler: An Efficient Patch-Independent Diffusion Model for Image Super-Resolution

Stability and Generalization of Zeroth-Order Decentralized Stochastic Gradient Descent with Changing Topology

GroundingFace: Fine-grained Face Understanding via Pixel Grounding Multimodal Large Language Model

Unsupervised Continual Anomaly Detection with Contrastively-Learned Prompt

ASWT-SGNN: Adaptive Spectral Wavelet Transform-Based Self-Supervised Graph Neural Network

WaveNet: Tackling Non-stationary Graph Signals via Graph Spectral Wavelets

Learning Multi-Scale Video-Text Correspondence for Weakly Supervised Temporal Article Gronding

Perfect Alignment May be Poisonous to Graph Contrastive Learning

Bridging the Gap: A Unified Video Comprehension Framework for Moment Retrieval and Highlight Detection

AdaO2B: Adaptive Online to Batch Conversion for Out-of-Distribution Generalization

MaxQ: Multi-Axis Query for N:M Sparsity Network

Open-Vocabulary Segmentation with Semantic-Assisted Calibration

VQA4CIR: Boosting Composed Image Retrieval with Visual Question Answering

An Aggregation-Free Federated Learning for Tackling Data Heterogeneity

SDSTrack: Self-Distillation Symmetric Adapter Learning for Multi-Modal Visual Object Tracking

Timer: Generative Pre-trained Transformers Are Large Time Series Models

Look Back for More: Harnessing Historical Sequential Updates for Personalized Federated Adapter Tuning

Driving in the Occupancy World: Vision-Centric 4D Occupancy Forecasting and Planning via World Models for Autonomous Driving

Concentration Inequalities for General Functions of Heavy-Tailed Random Variables

Algorithmic Stability Unleashed: Generalization Bounds with Unbounded Losses

IteRPrimE: Zero-shot Referring Image Segmentation with Iterative Grad-CAM Refinement and Primary Word Emphasis

LiT: Delving into a Simple Linear Diffusion Transformer for Image Generation