Research map / Vision-language and multimodal / Multimodal LLMs
Unveiling the Invisible: Reasoning Complex Occlusions Amodally with AURA
Zhixuan Li, Hyunse Yoon, Sanghoon Lee, Weisi Lin
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 Occlusion Condition Encoder (ablation) | AURA baseline · AmodalReasonSeg val | amodal gIoU: 43.55 -> 44.86 | Improved |
| add Spatial Occlusion Encoder (ablation) | AURA baseline · AmodalReasonSeg val | amodal gIoU: 43.55 -> 46.95 | Improved |
| add both occlusion encoders (ablation) | AURA baseline · AmodalReasonSeg val | amodal gIoU: 43.55 -> 47.76 | Improved |
| add both occlusion encoders (ablation) | AURA baseline · AmodalReasonSeg val | amodal cIoU: 43.92 -> 47.32 | Improved |
| add both occlusion encoders (ablation) | AURA baseline · AmodalReasonSeg val | visible gIoU: 48.96 -> 51.31 | Improved |
| add both occlusion encoders (ablation) | AURA baseline · AmodalReasonSeg val | visible cIoU: 48.63 -> 55.38 | Improved |
| add LoRA fine-tuning of the MLLM (ablation) | AURA · AmodalReasonSeg val | amodal gIoU: 20.84 -> 47.76 | Improved |
| add LoRA fine-tuning of the MLLM (ablation) | AURA · AmodalReasonSeg val | visible cIoU: 27.95 -> 55.38 | Improved |
| scale MLLM from LLaVA-7B to LLaVA-13B (ablation) | AURA · AmodalReasonSeg val | amodal gIoU: 47.76 -> 49.41 | Improved |
| upgrade MLLM to LLaVA-Llama2-13B (ablation) | AURA · AmodalReasonSeg val | amodal gIoU: 47.76 -> 52.34 | Improved |
| scale SAM backbone from ViT-B to ViT-H (ablation) | AURA · AmodalReasonSeg val | amodal gIoU: 42.85 -> 47.76 | Improved |
| scale SAM backbone from ViT-L to ViT-H (ablation) | AURA · AmodalReasonSeg val | amodal gIoU: 45.23 -> 47.76 | 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.
- GLaMM — GLaMM: Pixel Grounding Large Multimodal Model (CVPR 2024) · better on 2 of 2 shared results
- OMG-LLaVA — OMG-LLaVA: Bridging Image-level, Object-level, Pixel-level Reasoning and Understanding (NeurIPS 2024) · better on 2 of 2 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
- Open-World Amodal Appearance Completion (CVPR 2025 · similarity 0.88)
- Amodal Ground Truth and Completion in the Wild (CVPR 2024 · similarity 0.88)
- Amodal Scene Analysis via Holistic Occlusion Relation Inference and Generative Mask Completion (AAAI 2024 · similarity 0.88)
- Segment Anything, Even Occluded (CVPR 2025 · similarity 0.87)
- Using Diffusion Priors for Video Amodal Segmentation (CVPR 2025 · similarity 0.87)
- Towards Efficient Foundation Model for Zero-shot Amodal Segmentation (CVPR 2025 · 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.
