Research map / Graph learning / Graph algorithms and clustering
Graph Wasserstein Correlation Analysis for Movie Retrieval
Xueya Zhang, Tong Zhang, Xiaobin Hong, Zhen Cui, Jian Yang
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 |
|---|---|---|---|
| replace cosine with Wasserstein metric (ablation) | PCA features · MovieGraphs description retrieval | R@1: 35.9 -> 62.1 | Improved |
| replace cosine with Wasserstein metric (ablation) | CCA features · MovieGraphs description retrieval | R@1: 61.9 -> 66.4 | Improved |
| add joint graph signal filtering with Wasserstein metric learning (GWCA) over CCA with w-2 (ablation) | CCA w-2 · MovieGraphs description retrieval | R@5: 82.7 -> 87.8 | Improved |
| add learned projection to original features (ablation) | original features with cosine · MovieGraphs description retrieval | R@1: 4.07 -> 35.9 | Improved |
| use GWCA instead of GloVe max-sum pooling | description retrieval · MovieGraphs | R@1: 62.1 -> 67.6 | Improved |
| use max-sum pooling instead of idf-weighted max-sum | GloVe · MovieGraphs description retrieval | R@5: 81.6 -> 81.3 | No clear effect |
| add attribute nodes to scene and situation nodes (ablation) | video graph · MovieGraphs video retrieval | R@10: 8.7 -> 15.5 | Improved |
| add situation nodes to scene nodes (ablation) | video graph · MovieGraphs video retrieval | R@1: 1.1 -> 1.0 | No clear effect |
| use GWCA instead of cosine on sc,si,a graph nodes | video retrieval · MovieGraphs video retrieval | R@1: 2.2 -> 2.4 | Improved |
| set order of receptive field k to 2 with fusion (ablation) | GWCA · MovieGraphs description retrieval | R@5 | Improved |
| set node feature dimension to 240 (ablation) | GWCA · MovieGraphs description retrieval | R@5 | Improved |
Closest papers by meaning
- Wasserstein Coupled Graph Learning for Cross-Modal Retrieval (ICCV 2021 · similarity 0.85)
- Deep Wasserstein Graph Discriminant Learning for Graph Classification (AAAI 2021 · similarity 0.84)
- Wasserstein Weisfeiler-Lehman Graph Kernels (NeurIPS 2019 · similarity 0.84)
- Large Scale Video Representation Learning via Relational Graph Clustering (CVPR 2020 · similarity 0.84)
- Representing Graphs via Gromov-Wasserstein Factorization (TPAMI 2022 · similarity 0.84)
- Joint Hierarchical Representation Learning of Samples and Features via Informed Tree-Wasserstein Distance (NeurIPS 2025 · similarity 0.81)
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.
