Research map / Vision-language and multimodal / Multimodal LLMs
Perspective-Aware Reasoning in Vision-Language Models via Mental Imagery Simulation
Phillip Y. Lee, Jihyeon Je, Chanho Park, Mikaela Angelina Uy, Leonidas Guibas, Minhyuk Sung
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 |
|---|---|---|---|
| apply abstract perspective change with visual prompt | Qwen2.5-VL · COMFORT++ left/right | accuracy (%): 43.33 -> 89.67 | Improved |
| apply abstract perspective change with numerical prompt | Qwen2.5-VL · COMFORT++ left/right | accuracy (%): 43.33 -> 88.67 | Improved |
| apply abstract perspective change with visual prompt | Qwen2.5-VL · COMFORT++ facing | accuracy (%): 43.00 -> 88.33 | Improved |
| apply abstract perspective change with visual prompt | Qwen2.5-VL · 3DSRBench left/right | accuracy (%): 34.96 -> 72.78 | Improved |
| apply abstract perspective change with visual prompt | Qwen2.5-VL · 3DSRBench visibility | accuracy (%): 45.06 -> 67.44 | Improved |
| apply abstract perspective change with visual prompt | Qwen2.5-VL · 3DSRBench facing | accuracy (%): 53.47 -> 66.47 | Improved |
| use visual prompt instead of numerical prompt | APC · COMFORT++ visibility | accuracy (%): 71.25 -> 90.00 | Improved |
| use visual prompt instead of numerical prompt | APC · COMFORT++ facing | accuracy (%): 62.00 -> 88.33 | Improved |
| use numerical prompt instead of visual prompt | APC · COMFORT++ closer | accuracy (%): 94.33 -> 96.00 | Hurt |
| use abstraction-based APC instead of dense reconstruction ViewCrafter | ViewCrafter · 3DSRBench left/right | accuracy (%): 28.41 -> 72.78 | Improved |
| use APC abstraction instead of SpatialPIN* dense reconstruction | SpatialPIN* · COMFORT++ left/right | accuracy (%): 59.80 -> 89.67 | Improved |
| use APC instead of dense reconstruction in inference time | SpatialPIN* · Inference time per question | time (s): 336.21 -> 17.47 | Improved |
Closest papers by meaning
- SpinBench: Perspective and Rotation as a Lens on Spatial Reasoning in VLMs (ICLR 2026 · similarity 0.92)
- MMPerspective: Do MLLMs Understand Perspective? A Comprehensive Benchmark for Perspective Perception, Reasoning, and Robustness (NeurIPS 2025 · similarity 0.92)
- Thinking with Imagination: Agentic Visual Spatial Reasoning with World Simulators (NeurIPS 2026 · similarity 0.91)
- CoFFT: Chain of Foresight-Focus Thought for Visual Language Models (NeurIPS 2025 · similarity 0.90)
- 3ViewSense: Spatial and Mental Perspective Reasoning from Orthographic Views in Vision-Language Models (ICML 2026 · similarity 0.90)
- Machine Mental Imagery: Empower Multimodal Reasoning with Latent Visual Tokens (CVPR 2026 · similarity 0.90)
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.
