Research map / Label-efficient and robust learning / Recognition and long-tail
SPL-MLL: Selecting Predictable Landmarks for Multi-Label Learning
Junbing Li, Changqing Zhang, Pengfei Zhu, Baoyuan Wu, Lei Chen, Qinghua Hu
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 landmark selection and landmark prediction on top of NN embeddings (ablation) | NN-embeddings · pascal VOC 2007 | Average Precision: 0.431 -> 0.616 | Improved |
| add landmark selection and landmark prediction on top of NN embeddings (ablation) | NN-embeddings · pascal VOC 2007 | Micro-F1: 0.308 -> 0.586 | Improved |
| learn landmark selection and prediction jointly instead of separately (ablation) | Ours (NN + separated) · pascal VOC 2007 | Average Precision: 0.521 -> 0.616 | Improved |
| learn landmark selection and prediction jointly instead of separately (ablation) | Ours (NN + separated) · pascal VOC 2007 | Micro-F1: 0.384 -> 0.586 | Improved |
| learn landmark selection and prediction jointly instead of separately (ablation) | Ours (NN + separated) · pascal VOC 2007 | Ranking Loss: 0.243 -> 0.184 | Improved |
| use neural network embedding instead of linear projection (ablation) | Ours (joint + linear) · pascal VOC 2007 | Micro-F1: 0.516 -> 0.586 | Improved |
| use neural network embedding instead of linear projection (ablation) | Ours (joint + linear) · pascal VOC 2007 | Average Precision: 0.608 -> 0.616 | Improved |
| use neural network embedding instead of linear projection (ablation) | Ours (linear) · emotions | Average Precision: 0.798 -> 0.815 | Improved |
| use neural network embedding instead of linear projection (ablation) | Ours (linear) · scene | Average Precision: 0.861 -> 0.884 | Improved |
| use neural network embedding instead of linear projection (ablation) | Ours (linear) · tmc2007 | Micro-F1: 0.912 -> 0.944 | Improved |
| use neural network embedding instead of linear projection (ablation) | Ours (linear) · espgame | Average Precision: 0.289 -> 0.291 | No clear effect |
| add joint landmark selection over linear label-space baselines (2-step landmark methods) (ablation) | ML-CSSP · tmc2007 | Micro-F1: 0.604 -> 0.944 | Improved |
Closest papers by meaning
- Complementary to Multiple Labels: A Correlation-Aware Correction Approach (TPAMI 2024 · similarity 0.85)
- Label Distribution Learning by Exploiting Label Correlations (AAAI 2018 · similarity 0.85)
- Multi-Label Learning With Multiple Complementary Labels (TPAMI 2025 · similarity 0.85)
- Realistic Evaluation of Deep Partial-Label Learning Algorithms (ICLR 2025 · similarity 0.85)
- Double-Layer Hybrid-Label Identification Feature Selection for Multi-View Multi-Label Learning (AAAI 2024 · similarity 0.84)
- Addressing Multi-Label Learning with Partial Labels: From Sample Selection to Label Selection (AAAI 2025 · 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.
