Research map / Vision-language and multimodal / Visual question answering
3D-Aware Visual Question Answering about Parts, Poses and Occlusions
Xingrui Wang, Wufei Ma, Zhuowan Li, Adam Kortylewski, Alan Yuille
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 |
|---|---|---|---|
| Use mesh projection to locate parts and occlusions instead of 2D part boxes (ablation) | PNSVQA · Super-CLEVR-3D test (all question types) | accuracy: 64.39 -> 68.15 | Improved |
| Use mesh projection to locate parts and occlusions instead of 2D part boxes (ablation) | PNSVQA · Super-CLEVR-3D test (part questions) | accuracy: 50.61 -> 56.30 | Improved |
| Use mesh projection to locate parts and occlusions instead of 2D part boxes (ablation) | PNSVQA · Super-CLEVR-3D test (pose questions) | accuracy: 87.78 -> 86.70 | Hurt |
| Use mesh projection to locate parts and occlusions instead of 2D part boxes (ablation) | PNSVQA · Super-CLEVR-3D test (occlusion questions) | accuracy: 65.80 -> 70.70 | Improved |
| Replace regression-based pose prediction with generative render-and-compare 6D pose estimation (ablation) | PNSVQA+Projection · Super-CLEVR-3D test (all question types) | accuracy: 68.15 -> 75.64 | Improved |
| Replace regression-based pose prediction with generative render-and-compare 6D pose estimation (ablation) | PNSVQA+Projection · Super-CLEVR-3D test (part questions) | accuracy: 56.30 -> 71.85 | Improved |
| Replace regression-based pose prediction with generative render-and-compare 6D pose estimation (ablation) | PNSVQA+Projection · Super-CLEVR-3D test (pose questions) | accuracy: 86.70 -> 86.40 | Hurt |
| Replace regression-based pose prediction with generative render-and-compare 6D pose estimation (ablation) | PNSVQA+Projection · Super-CLEVR-3D test (occlusion questions) | accuracy: 70.70 -> 76.90 | Improved |
| Replace regression-based pose prediction with generative render-and-compare 6D pose estimation (ablation) | PNSVQA+Projection · Super-CLEVR-3D test (part+occlusion questions) | accuracy: 58.90 -> 67.40 | Improved |
| Use PO3D-VQA's 3D-aware scene parsing and probabilistic modular reasoning instead of PNSVQA's 2D parsing | PNSVQA · Super-CLEVR-3D test (all question types) | accuracy: 64.39 -> 75.64 | Improved |
| Use PO3D-VQA instead of PNSVQA | PNSVQA · Super-CLEVR-3D test (pose questions) | accuracy: 87.78 -> 86.40 | Hurt |
| Use PO3D-VQA instead of PNSVQA | PNSVQA · Super-CLEVR-3D test (part questions) | accuracy: 50.61 -> 71.85 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- FiLM — FILM: Following Instructions in Language with Modular Methods (ICLR 2022) · better on 5 of 5 shared results
- mDETR — MDETR - Modulated Detection for End-to-End Multi-Modal Understanding (ICCV 2021) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Bridging the Gap between 2D and 3D Visual Question Answering: A Fusion Approach for 3D VQA (AAAI 2024 · similarity 0.93)
- Beyond 3D VQAs: Injecting 3D Spatial Priors into Vision-Language Models for Enhanced Geometric Reasoning (CVPR 2026 · similarity 0.90)
- 3D Question Answering via only 2D Vision-Language Models (ICML 2025 · similarity 0.89)
- Unveiling the Mist over 3D Vision-Language Understanding: Object-centric Evaluation with Chain-of-Analysis (CVPR 2025 · similarity 0.89)
- When Open-Vocabulary Visual Question Answering Meets Causal Adapter: Benchmark and Approach (AAAI 2025 · similarity 0.89)
- VLM-3R: Vision-Language Models Augmented with Instruction-Aligned 3D Reconstruction (CVPR 2026 · similarity 0.88)
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.
