Pay More Attention To Text In High-Resolution MLLMs
cs.CV
Submitted: 2026-09-20
Updated: 2026-09-20
Terminology
Sources
- Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond
- Accelerating Multimodal Large Language Models with Prior-Corrected Token Reduction
- How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites
- InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks
- PaLM-E: An Embodied Multimodal Language Model
- GPT-4o System Card
- DeepScan: A Training-Free Framework for Visually Grounded Reasoning in Large Vision-Language Models
- HiDe: Rethinking The Zoom-IN method in High Resolution MLLMs via Hierarchical Decoupling
- Thinking Once Is Enough: Intermediate-Layer Evidence Routing for High-Resolution VQA
- InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency
- Zooming without Zooming: Region-to-Image Distillation for Fine-Grained Multimodal Perception
- Vision-OPD: Learning to See Fine Details for Multimodal LLMs via On-Policy Self-Distillation
- Samba+: General and Accurate Salient Object Detection via A More Unified Mamba-based Framework
- Is It Time for the Renaissance of Salient Object Detection in the Era of MLLMs?
- InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models
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