IoU-PD: IoU-Aware Privileged Distillation for Visual Grounding with Multimodal Large Language Models

arXiv:2607.15732 · cs.CV · Submitted 2026-07-17 · Read on arXiv

cs.CV

Submitted: 2026-07-17

Updated: 2026-10-13

Comments: 16 pages, 7 figures, 11 tables

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Grounding — 4× in this paper · explained in 3 episodes
• Distillation — 3× in this paper · explained in 2 episodes
• Visual Grounding — 2× in this paper · explained in 2 episodes
• Benchmark — 1× in this paper · explained in 2 episodes
• Fine-Tuning — 1× in this paper · explained in 5 episodes

Sources

• Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond
• Qwen3-VL Technical Report
• Qwen2.5-VL Technical Report
• Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic
• Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling
• How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites
• Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities
• Seed1.5-VL Technical Report
• Distilling the Knowledge in a Neural Network
• GLM-5V-Turbo: Toward a Native Foundation Model for Multimodal Agents
• GLM-4.5V and GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning
• Detect Anything via Next Point Prediction
• Smooth Operator: Smooth Verifiable Reward Activates Spatial Reasoning Ability of Vision-Language Model
• Visual-RFT: Visual Reinforcement Fine-Tuning
• Kosmos-2: Grounding Multimodal Large Language Models to the World
• Seed1.8 Model Card: Towards Generalized Real-World Agency
• Seed2.0 Model Card: Towards Intelligence Frontier for Real-World Complexity
• VLM-R1: A Stable and Generalizable R1-style Large Vision-Language Model
• Enhancing the Reasoning Ability of Multimodal Large Language Models via Mixed Preference Optimization
• DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models