GeoScope: Training-Free Program Execution beyond the Counting Horizon for Ultra-High-Resolution Remote Sensing Question Answering
cs.CV, cs.AI
Submitted: 2026-11-09
Updated: 2026-11-09
Project page: https://avalon-s.github.io/GeoScope
Terminology
Sources
- Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAs
- More with Less: a Large Scale Remote Sensing VLM with a Simple Recipe
- Getting to the Point: Pointing Improves LVLMs at Counting
- CounterCount: Diagnosing Knowledge Bias in How Vision-Language Models Count Attributes
- Qwen3-VL Technical Report
- Qwen2.5-VL Technical Report
- Why MLLMs Struggle to Count: Overcoming Individuation and Aggregation Bottlenecks with ConvStack
- A Benchmark for Ultra-High-Resolution Remote Sensing MLLMs
- The Llama 3 Herd of Models
- Your Vision-Language Model Can't Even Count to 20: Exposing the Failures of VLMs in Compositional Counting
- Geode: A Zero-shot Geospatial Question-Answering Agent with Explicit Reasoning and Precise Spatio-Temporal Retrieval
- TerraAgent: Demonstration-Grounded Multi-Agent Reasoning for Long-Context Remote Sensing Visual Question Answering
- RS-OPSD: Reliable Privileged On-Policy-Self-Distillation for Ultra-High-Resolution Remote Sensing VQA
- Seeing Clearly without Training: Mitigating Hallucinations in Multimodal LLMs for Remote Sensing
- SkySenseGPT: A Fine-Grained Instruction Tuning Dataset and Model for Remote Sensing Vision-Language Understanding
- WeaveEarth: Structured Evidence Construction and Reasoning for Training-Free UHR Remote Sensing Understanding
- SQuID: A Benchmark for Quantitative Spatial Reasoning on Satellite Imagery
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs