Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
KG-SP: Knowledge Guided Simple Primitives for Open World Compositional Zero-Shot Learning
Shyamgopal Karthik, Massimiliano Mancini, Zeynep Akata
Check 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 |
|---|---|---|---|
| predict states and objects separately rather than directly over compositions (ablation) | CGE · MIT-States validation OW-CZSL | HM: 7.0 -> 8.1 | Improved |
| add ConceptNet-based hard masking of infeasible compositions (ablation) | KG-SP-ff · MIT-States validation OW-CZSL | Unseen accuracy: 7.4 -> 7.7 | Improved |
| add ConceptNet-guided pseudo-labeling (ablation) | KG-SP · MIT-States validation pCZSL | HM: 4.8 -> 5.3 | Improved |
| replace knowledge-guided pseudo-labeling with vanilla pseudo-labeling (ablation) | KG-SP · MIT-States validation pCZSL | Seen accuracy: 16.6 vs 15.9 | Hurt |
| use Gumbel-softmax sampling without external feasibility knowledge (ablation) | KG-SP · MIT-States validation pCZSL | HM: 4.8 vs 4.5 | Hurt |
| increase classifier depth from 3 to 5 layers (ablation) | KG-SP-ff · UT-Zappos validation OW-CZSL | AUC: 26.9 to 24.3 | Hurt |
| add entropy minimization (ablation) | CGE · C-GQA pCZSL | AUC: 1.4 -> 1.8 | Improved |
| add entropy minimization (ablation) | CGE · UT-Zappos pCZSL | AUC: 10.7 -> 8.7 | Hurt |
| add pseudo-labeling (ablation) | CompCos · MIT-States pCZSL | AUC: 3.6 to 3.2 | Hurt |
| use ConceptNet feasibility scores instead of no masking (ablation) | KG-SP-ff · MIT-States validation OW-CZSL | HM: 7.9 -> 8.2 | Improved |
| use deeper nonlinear primitive classifiers instead of shallow predictors (ablation) | KG-SP-ff · UT-Zappos and MIT-States validation OW-CZSL | AUC | Improved |
Later papers that compared with this method
2 later accepted papers report KG-SP as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Simple Primitives With Feasibility- and Contextuality-Dependence for Open-World Compositional Zero-Shot Learning (TPAMI 2023 · better on 6 of 6 shared results)
- Distilled Reverse Attention Network for Open-world Compositional Zero-Shot Learning (ICCV 2023 · 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
- ProCC: Progressive Cross-Primitive Compatibility for Open-World Compositional Zero-Shot Learning (AAAI 2024 · similarity 0.92)
- Retrieval-Augmented Primitive Representations for Compositional Zero-Shot Learning (AAAI 2024 · similarity 0.92)
- A Dynamic Learning Method towards Realistic Compositional Zero-Shot Learning (AAAI 2024 · similarity 0.91)
- Context-based and Diversity-driven Specificity in Compositional Zero-Shot Learning (CVPR 2024 · similarity 0.91)
- Beyond Seen Primitive Concepts and Attribute-Object Compositional Learning (CVPR 2024 · similarity 0.90)
- Troika: Multi-Path Cross-Modal Traction for Compositional Zero-Shot Learning (CVPR 2024 · similarity 0.90)
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.
