Research map / 3D vision / 3D generation
A Deep Learning Model of Mental Rotation Informed by Interactive VR Experiments
Raymond Khazoum, Daniela Fernandes, Aleksandr Krylov, Qin Li, Stephane Deny
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 Module I equivariant 3D spatial representation before the symbolic encoder (ablation) | mental rotation model · Mental rotation | task solvable | Improved |
| add VSM symbolic module with autoregressive decoder instead of frozen 3D conv (ablation) | mental rotation model · Mental rotation | accuracy (%): 38.46 -> 96.13 | Improved |
| add VSM autoregressive decoder on top of ViT encoder (ablation) | mental rotation model · Mental rotation | accuracy (%): 90.71 -> 96.13 | Improved |
| add VSM autoregressive decoder vs patch embedding only (ablation) | mental rotation model · Mental rotation | accuracy (%): 36.14 -> 96.13 | Improved |
| add rotation actions to Module III instead of direct similarity decision (ablation) | mental rotation model · Mental rotation | accuracy (%): 97.03 -> 96.13 | No clear effect |
| add rotation actions with a one-layer MLP (ablation) | one-layer MLP · Mental rotation | accuracy (%): 86.23 -> 96.05 | Improved |
| use Siamese feed-forward encoders instead of equivariant-symbolic pipeline (ablation) | ResNet / ViT · Mental rotation test objects | accuracy: ~50% chance | Hurt |
| let humans manipulate shapes (Action setup) vs No-Action | human subjects · VR mental rotation | accuracy (%): 91.14 -> 95.33 | Improved |
| let humans manipulate shapes (Action setup) | human subjects · VR mental rotation | mean RT match (s): 3.72 -> 3.40 | Improved |
Closest papers by meaning
- SpinBench: Perspective and Rotation as a Lens on Spatial Reasoning in VLMs (ICLR 2026 · similarity 0.87)
- Spatial-IQ: Deconstructing Spatial Intelligence via Hierarchical Capability Tests (NeurIPS 2026 · similarity 0.87)
- MindJourney: Test-Time Scaling with World Models for Spatial Reasoning (NeurIPS 2025 · similarity 0.86)
- 3ViewSense: Spatial and Mental Perspective Reasoning from Orthographic Views in Vision-Language Models (ICML 2026 · similarity 0.86)
- Unfolding Spatial Cognition: Evaluating Multimodal Models on Visual Simulations (ICLR 2026 · similarity 0.85)
- InternSpatial: A Comprehensive Dataset for Spatial Reasoning in Vision-Language Models (ICLR 2026 · similarity 0.85)
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.
