Research map / Vision-language and multimodal / Human and embodied modeling
Exploring Figure-Ground Assignment Mechanism in Perceptual Organization
Wei Zhai, Yang Cao, Jing Zhang, Zheng-Jun Zha
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 LR and convexity cues plus IEM to UNet baseline (full FGA-Net) | UNet · Figure-Ground Segregation normal | S-measure: .745 -> .816 | Improved |
| full FGA-Net vs baseline | UNet · Figure-Ground Segregation normal | IoU: .580 -> .659 | Improved |
| remove IEM from FGA-Net (LR + Convexity only) (ablation) | FGA-Net · Figure-Ground Segregation normal | S-measure: .816 -> .793 | Hurt |
| add only a single cue to baseline (ablation) | UNet · Figure-Ground Segregation normal | S-measure: .745 -> .771 | Improved |
| use two branches for both priors rather than a single branch (ablation) | FGA-Net · Figure-Ground Segregation | performance | Improved |
| add collaborative strategy to local interaction in IEM (ablation) | IEM · Figure-Ground Segregation normal | S-measure: .801 -> .805 | Improved |
| add Lambda strategy to IEM (full IEM) (ablation) | IEM · Figure-Ground Segregation normal | S-measure: .810 -> .816 | Improved |
| local interaction vs global interaction in IEM (ablation) | IEM · Figure-Ground Segregation normal | S-measure: .797 -> .801 | Improved |
| FGA-Net vs adding direct edge supervision | UNet · Figure-Ground Segregation | segmentation completeness | Improved |
| training on only 1% of the data with FGA | FGA-Net · Figure-Ground Segregation | performance | No clear effect |
| FGA-Net vs SINetv2 on COD10K | FGA-Net · COD10K | MAE: .037 -> .031 | Improved |
| FGA-Net vs second-best on ETIS Dice | FGA-Net · ETIS | Dice: .719 -> .723 | Improved |
Closest papers by meaning
- Vision Transformers Learn Gestalt-Like Figure-Ground Cues from Natural Images (NeurIPS 2026 · similarity 0.90)
- L-WISE: Boosting Human Visual Category Learning Through Model-Based Image Selection and Enhancement (ICLR 2025 · similarity 0.85)
- When does perceptual alignment benefit vision representations? (NeurIPS 2024 · similarity 0.85)
- Mind the GAP: Glimpse-based Active Perception improves generalization and sample efficiency of visual reasoning (ICLR 2025 · similarity 0.85)
- Perceptual Group Tokenizer: Building Perception with Iterative Grouping (ICLR 2024 · similarity 0.85)
- Beyond the Doors of Perception: Vision Transformers Represent Relations Between Objects (NeurIPS 2024 · similarity 0.84)
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.
