Cover Image

Semantic-Geometric Dual Compression: Training-Free Visual Token Reduction for Ultra-High-Resolution Remote Sensing Understanding

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Li, Yueying, Wang, Fengxiang, Li, Yan, Chen, Mingshuo, Zhao, Mengying, Lan, Long
Format:	Preprint
Published:	2026
Subjects:	Computer Vision and Pattern Recognition Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

GeoEyes: On-Demand Visual Focusing for Evidence-Grounded Understanding of Ultra-High-Resolution Remote Sensing Imagery
by: Wang, Fengxiang, et al.
Published: (2026)

XLRS-Bench: Could Your Multimodal LLMs Understand Extremely Large Ultra-High-Resolution Remote Sensing Imagery?
by: Wang, Fengxiang, et al.
Published: (2025)

GeoLLaVA-8K: Scaling Remote-Sensing Multimodal Large Language Models to 8K Resolution
by: Wang, Fengxiang, et al.
Published: (2025)

Text Before Vision: Staged Knowledge Injection Matters for Agentic RLVR in Ultra-High-Resolution Remote Sensing Understanding
by: Wang, Fengxiang, et al.
Published: (2026)

UHR-BAT: Budget-Aware Token Compression Vision-Language model for Ultra-High-Resolution Remote Sensing
by: Dang, Yunkai, et al.
Published: (2026)

RoMA: Scaling up Mamba-based Foundation Models for Remote Sensing
by: Wang, Fengxiang, et al.
Published: (2025)

GeoVista: Visually Grounded Active Perception for Ultra-High-Resolution Remote Sensing Understanding
by: Zhu, Jiashun, et al.
Published: (2026)

Homogeneous Tokenizer Matters: Homogeneous Visual Tokenizer for Remote Sensing Image Understanding
by: Shao, Run, et al.
Published: (2024)

Filter, Correlate, Compress: Training-Free Token Reduction for MLLM Acceleration
by: Han, Yuhang, et al.
Published: (2024)

A Benchmark for Ultra-High-Resolution Remote Sensing MLLMs
by: Dang, Yunkai, et al.
Published: (2025)

Training-Free Point Cloud Recognition Based on Geometric and Semantic Information Fusion
by: Chen, Yan, et al.
Published: (2024)

DualToken: Towards Unifying Visual Understanding and Generation with Dual Visual Vocabularies
by: Song, Wei, et al.
Published: (2025)

Dual-Branch Remote Sensing Infrared Image Super-Resolution
by: Ge, Xining, et al.
Published: (2026)

InstructSAM: A Training-Free Framework for Instruction-Oriented Remote Sensing Object Recognition
by: Zheng, Yijie, et al.
Published: (2025)

GeoViS: Geospatially Rewarded Visual Search for Remote Sensing Visual Grounding
by: Zhang, Peirong, et al.
Published: (2025)

Look Where It Matters: Training-Free Ultra-HR Remote Sensing VQA via Adaptive Zoom Search
by: Zhou, Yunqi, et al.
Published: (2025)

Adapting Vision Transformers to Ultra-High Resolution Semantic Segmentation with Relay Tokens
by: Perron, Yohann, et al.
Published: (2026)

SeGPruner: Semantic-Geometric Visual Token Pruner for 3D Question Answering
by: Li, Wenli, et al.
Published: (2026)

ImageRAG: Enhancing Ultra High Resolution Remote Sensing Imagery Analysis with ImageRAG
by: Zhang, Zilun, et al.
Published: (2024)

Co-Training Vision Language Models for Remote Sensing Multi-task Learning
by: Li, Qingyun, et al.
Published: (2025)

LMFNet: An Efficient Multimodal Fusion Approach for Semantic Segmentation in High-Resolution Remote Sensing
by: Wang, Tong, et al.
Published: (2024)

UHR-DETR: Efficient End-to-End Small Object Detection for Ultra-High-Resolution Remote Sensing Imagery
by: Li, Jingfang, et al.
Published: (2026)

Geometric Knowledge-Assisted Federated Dual Knowledge Distillation Approach Towards Remote Sensing Satellite Imagery
by: Zou, Luyao, et al.
Published: (2026)

DGL-RSIS: Decoupling Global Spatial Context and Local Class Semantics for Training-Free Remote Sensing Image Segmentation
by: Li, Boyi, et al.
Published: (2025)

Highly Compressed Tokenizer Can Generate Without Training
by: Beyer, L. Lao, et al.
Published: (2025)

Structure-Semantic Decoupled Modulation of Global Geospatial Embeddings for High-Resolution Remote Sensing Mapping
by: Lyu, Jienan, et al.
Published: (2026)

Reconciling Semantic Controllability and Diversity for Remote Sensing Image Synthesis with Hybrid Semantic Embedding
by: Liu, Junde, et al.
Published: (2024)

GeoSeg: Training-Free Reasoning-Driven Segmentation in Remote Sensing Imagery
by: Jiang, Lifan, et al.
Published: (2026)

Text2Seg: Remote Sensing Image Semantic Segmentation via Text-Guided Visual Foundation Models
by: Zhang, Jielu, et al.
Published: (2023)

Video-XL-Pro: Reconstructive Token Compression for Extremely Long Video Understanding
by: Liu, Xiangrui, et al.
Published: (2025)

A Global-Local Cross-Attention Network for Ultra-high Resolution Remote Sensing Image Semantic Segmentation
by: Yi, Chen, et al.
Published: (2025)

Multilingual Training-Free Remote Sensing Image Captioning
by: Rebelo, Carlos, et al.
Published: (2025)

VisionTrim: Unified Vision Token Compression for Training-Free MLLM Acceleration
by: Yu, Hanxun, et al.
Published: (2026)

F2Net: A Frequency-Fused Network for Ultra-High Resolution Remote Sensing Segmentation
by: Chen, Hengzhi, et al.
Published: (2025)

Semantically Robust Unsupervised Image Translation for Paired Remote Sensing Images
by: Fang, Sheng, et al.
Published: (2025)

Ultra-Low Complexity On-Orbit Compression for Remote Sensing Imagery via Block Modulated Imaging
by: Wang, Zhibin, et al.
Published: (2024)

UNetMamba: An Efficient UNet-Like Mamba for Semantic Segmentation of High-Resolution Remote Sensing Images
by: Zhu, Enze, et al.
Published: (2024)

Spatial-Regularization-Aware Dual-Branch Collaborative Inference for Training-Free OVSS in Remote Sensing Imagery
by: Wang, Jianzheng, et al.
Published: (2026)

UniCompress: Token Compression for Unified Vision-Language Understanding and Generation
by: Wang, Ziyao, et al.
Published: (2026)

LFTR: Learning-Free Token Reduction for Multimodal Large Language Models
by: Zhao, Zihui, et al.
Published: (2025)