Research map / Detection and segmentation / CNN features
Affinity CNN: Learning Pixel-Centric Pairwise Relations for Figure/Ground Embedding
Michael Maire, Takuya Narihira, Stella X. Yu
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 Maire's figure/ground method with Affinity CNN | BSDS figure/ground predictions evaluated on ground-truth segmentation · BSDS | R-ACC: 0.56 -> 0.62 | Improved |
| Replace Maire's figure/ground method with Affinity CNN | BSDS figure/ground predictions evaluated on ground-truth segmentation · BSDS | B-ACC: 0.58 -> 0.69 | Improved |
| Replace Maire's figure/ground method with Affinity CNN | BSDS figure/ground predictions evaluated on ground-truth segmentation · BSDS | B-ACC-50: 0.56 -> 0.72 | Improved |
| Replace Maire's figure/ground method with Affinity CNN | BSDS figure/ground predictions evaluated on ground-truth segmentation · BSDS | B-ACC-25: 0.56 -> 0.73 | Improved |
| Replace Maire's figure/ground method with Affinity CNN | BSDS figure/ground predictions evaluated on Affinity CNN segmentation · BSDS | R-ACC: 0.59 -> 0.66 | Improved |
| Replace Maire's figure/ground method with Affinity CNN | BSDS figure/ground predictions evaluated on Affinity CNN segmentation · BSDS | B-ACC: 0.62 -> 0.70 | Improved |
| Replace Maire's figure/ground method with Affinity CNN | BSDS figure/ground predictions evaluated on Affinity CNN segmentation · BSDS | B-ACC-50: 0.61 -> 0.69 | Improved |
| Replace Maire's figure/ground method with Affinity CNN | BSDS figure/ground predictions evaluated on Affinity CNN segmentation · BSDS | B-ACC-25: 0.58 -> 0.67 | Improved |
| Use Affinity CNN spectral boundaries instead of the gPb spectral component | BSDS boundary detection · BSDS | F-measure: 0.68 -> 0.68 | No clear effect |
Closest papers by meaning
- Deciphering 'What' and 'Where' Visual Pathways from Spectral Clustering of Layer-Distributed Neural Representations (CVPR 2024 · similarity 0.88)
- Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised Semantic Segmentation and Localization (CVPR 2022 · similarity 0.86)
- Scene Categorization With Spectral Features (ICCV 2017 · similarity 0.86)
- Learning to Model Pixel-Embedded Affinity for Homogeneous Instance Segmentation (AAAI 2022 · similarity 0.86)
- The Unreasonable Effectiveness of Linear Prediction as a Perceptual Metric (ICLR 2024 · similarity 0.85)
- Vision Transformers Learn Gestalt-Like Figure-Ground Cues from Natural Images (NeurIPS 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.
