Research map / Detection and segmentation / Open-set and few-shot detection
SIRI: Spatial Relation Induced Network For Spatial Description Resolution
peiyao wang, Weixin Luo, Yanyu Xu, Haojie Li, Shugong Xu, Jianyu Yang, Shenghua Gao
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add the full SIRI network | LingUnet · Touchdown testing | A@80px (%): 34.59 -> 58.33 | Improved |
| Add the full SIRI network | LingUnet · Extended Touchdown | A@80px (%): 30.72 -> 48.96 | Improved |
| Add object-level visual feature correlation (ablation) | LingUnet · Touchdown testing | A@80px (%): 34.59 -> 37.42 | Improved |
| Add local spatial relationship guided distillation (ablation) | LingUnet · Touchdown testing | A@80px (%): 34.59 -> 44.95 | Improved |
| Add global spatial positional embedding (ablation) | LingUnet · Touchdown testing | A@80px (%): 34.59 -> 38.76 | Improved |
| Add object-level correlation and local relationship distillation (ablation) | LingUnet · Touchdown testing | A@80px (%): 34.59 -> 45.63 | Improved |
| Add local relationship distillation and global positional embedding (ablation) | LingUnet · Touchdown testing | A@80px (%): 34.59 -> 56.24 | Improved |
| Add object-level correlation, local relationship distillation, and global positional embedding (ablation) | LingUnet · Touchdown testing | A@80px (%): 34.59 -> 58.33 | Improved |
| Use 4 rather than 6 high-frequency orientation words (ablation) | SIRI · Touchdown testing | A@80px (%): 51.02 -> 58.33 | Hurt |
| Use 8 rather than 6 high-frequency orientation words (ablation) | SIRI · Touchdown testing | A@80px (%): 58.33 -> 59.74 | Improved |
| Replace LingUnet with convolutional layers | SIRI · Touchdown testing | A@80px (%): 58.33 -> 55.73 | Hurt |
| Introduce copy-paste visual ambiguities | SIRI · Touchdown testing | A@80px (%): 58.33 -> 52.86 | Hurt |
Closest papers by meaning
- Stitch and Tell: A Structured Data Augmentation Method for Spatial Understanding (NeurIPS 2025 · similarity 0.87)
- ERGeoBench: A Comprehensive Benchmark for Embodied Reasoning and Geo-localization in Multimodal Large Language Models (ICML 2026 · similarity 0.87)
- GRE Suite: Geo-localization Inference via Fine-Tuned Vision-Language Models and Enhanced Reasoning Chains (NeurIPS 2025 · similarity 0.86)
- Mechanisms of Object Localization in Vision-Language Models (CVPR 2026 · similarity 0.86)
- Where Am I and What Will I See: An Auto-Regressive Model for Spatial Localization and View Prediction (ICLR 2025 · similarity 0.86)
- SpatialRGPT: Grounded Spatial Reasoning in Vision-Language Models (NeurIPS 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
