Research map / Vision-language and multimodal / Human and embodied modeling
MAAL: Multimodality-Aware Autoencoder-Based Affordance Learning for 3D Articulated Objects
Yuanzhi Liang, Xiaohan Wang, Linchao Zhu, Yi Yang
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 |
|---|---|---|---|
| replace Where2Act and AdaAfford with MAAL | MAAL · PartNet-Mobility Pushing All (train cat.) | F-score: 73.21 -> 76.63 | Improved |
| MAAL versus AdaAfford on unseen test categories for pulling | MAAL · Pulling All (test cat.) | Sample-Succ: 6.02 -> 6.67 | Improved |
| add interaction branch with intermediate fusion to action and object branches (ablation) | MME · Pushing All (train cat.) | F-score: 59.87 -> 76.63 | Improved |
| add action branch to object plus interaction branches (ablation) | MME · Pushing All (train cat.) | F-score: 75.54 -> 76.63 | Improved |
| use intermediate fusion instead of early fusion in encoder (ablation) | MME · Pushing All (train cat.) | F-score: decrease of 8.31% in F-score for early fusion | Improved |
| use MAAL scoring on AdaAfford proposals (ablation) | AdaAfford proposals · Pushing All (train cat.) Sample-Succ | Sample-Succ: 32.50 -> 32.36 | No clear effect |
| train only on positive samples with an autoencoder in one stage | MAAL · data usage compared to AdaAfford | positive training data share: 17.69% of AdaAfford data for pushing | Improved |
Closest papers by meaning
- IAAO: Interactive Affordance Learning for Articulated Objects in 3D Environments (CVPR 2025 · similarity 0.90)
- GEAL: Generalizable 3D Affordance Learning with Cross-Modal Consistency (CVPR 2025 · similarity 0.90)
- AffordBot: 3D Fine-grained Embodied Reasoning via Multimodal Large Language Models (NeurIPS 2025 · similarity 0.89)
- Learning 2D Invariant Affordance Knowledge for 3D Affordance Grounding (AAAI 2025 · similarity 0.88)
- SeqAfford: Sequential 3D Affordance Reasoning via Multimodal Large Language Model (CVPR 2025 · similarity 0.88)
- Weakly Supervised Multimodal Affordance Grounding for Egocentric Images (AAAI 2024 · similarity 0.87)
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.
