Jun Yu

PointRePar : SpatioTemporal Point Relation Parsing for Robust Category-Unified 3D Tracking

Prune Redundancy, Preserve Essence: Vision Token Compression in VLMs via Synergistic Importance-Diversity

KaLM-Embedding-V2: Superior Training Techniques and Data Inspire A Versatile Embedding Model

DiffTrans: Differentiable Geometry-Materials Decomposition for Reconstructing Transparent Objects

Uni-X: Mitigating Modality Conflict with a Two-End-Separated Architecture for Unified Multimodal Models

Evaluating and Steering Modality Preference in Multimodal Large Language Model

Video-ToC: Video Tree-of-Cue Reasoning

DPA-SGG: Dual Prompt Learning with Pseudo-Visual Augmentation for Open-Vocabulary Scene Graph Generation

CoMo: Compositional Motion Customization for Text-to-Video Generation

Amodal SAM: Open-World Amodal Segmentation

From Bias to Balance: Exploring and Mitigating Spatial Bias in LVLMs

A Token is Worth over 1,000 Tokens: Efficient Knowledge Distillation through Low-Rank Clone

EditInfinity: Image Editing with Binary-Quantized Generative Models

Towards Generalizable Detector for Generated Image

Enhancing LLM Planning for Robotics Manipulation through Hierarchical Procedural Knowledge Graphs

CoVT-CXR: Building Chain of Visual Thought for Interpretable Chest X-Ray Diagnosis

Function-to-Style Guidance of LLMs for Code Translation

A General Framework for Producing Interpretable Semantic Text Embeddings

OmniKV: Dynamic Context Selection for Efficient Long-Context LLMs

Deep Kernel Relative Test for Machine-generated Text Detection

Enhancing Target-unspecific Tasks through a Features Matrix

Optimizing Preference Alignment with Differentiable NDCG Ranking