Research map / Graph learning / Graph self-supervised learning
PRISM: Partial-label Relational Inference with Spatial and Spectral Cues
Yiyang Gu, Wenrui Wu, Yifang Qin, Taian Guo, Tao Zhe, Jiaru Tang, Zhiping Xiao, Weizhi Zhang, Ziyue Qiao, Wei Ju, Dongjie Wang, Xiao Luo and 2 more
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 relational inference with label propagation (ablation) | PRISM · ENZYMES q=0.3 | accuracy: 57.78 -> 63.11 | Improved |
| add relational inference with label propagation (ablation) | PRISM · Letter-High q=0.5 | accuracy: 71.24 -> 78.32 | Improved |
| add relational inference with label propagation (ablation) | PRISM · CIFAR10 q=0.5 | accuracy: 49.79 -> 55.10 | Improved |
| add class-specific substructure alignment instead of holistic embeddings (ablation) | PRISM · ENZYMES q=0.5 | accuracy: 49.56 -> 51.33 | Improved |
| add class-specific substructure alignment (ablation) | PRISM · Letter-High q=0.5 | accuracy: 76.11 -> 78.32 | Improved |
| add spatial branch (ablation) | PRISM · CIFAR10 q=0.5 | accuracy: 51.72 -> 55.10 | Improved |
| add spatial branch (ablation) | PRISM · ENZYMES q=0.3 | accuracy: 60.89 -> 63.11 | Improved |
| add spectral multi-band attention (ablation) | PRISM · ENZYMES q=0.5 | accuracy: 48.89 -> 51.33 | Improved |
| add spectral multi-band attention (ablation) | PRISM · Letter-High q=0.3 | accuracy: 80.17 -> 82.55 | Improved |
| PRISM vs DEER on COIL-DEL at higher ambiguity | PRISM · COIL-DEL q=0.1 | accuracy: 68.03 -> 79.48 | Improved |
| PRISM vs second best on ENZYMES | PRISM · ENZYMES q=0.3 | accuracy: 58.22 -> 63.11 | Improved |
| increase top-ka and top-ke neighbors from 1 to larger values (ablation) | PRISM · ENZYMES / Letter-High | accuracy | Mixed |
Closest papers by meaning
- Optimal Block-wise Asymmetric Graph Construction for Graph-based Semi-supervised Learning (NeurIPS 2023 · similarity 0.88)
- Graph inference learning for semi-supervised classification (ICLR 2020 · similarity 0.88)
- Learning Semantic-Specific Graph Representation for Multi-Label Image Recognition (ICCV 2019 · similarity 0.87)
- Why Propagate Alone? Parallel Use of Labels and Features on Graphs (ICLR 2022 · similarity 0.86)
- A Flexible Generative Framework for Graph-based Semi-supervised Learning (NeurIPS 2019 · similarity 0.86)
- Adaptive Graph Guided Disambiguation for Partial Label Learning (TPAMI 2021 · 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.
