Research map / Continual learning / Class-incremental
Incremental Embedding Learning via Zero-Shot Translation
Kun Wei, Cheng Deng, Xu Yang, Maosen Li
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 zero-shot translation (ablation) | E-FT · CUB | average incremental accuracy: 38.7 -> 54.0 | Improved |
| Add zero-shot translation and unified representation (ablation) | E-FT · CUB | average incremental accuracy: 38.7 -> 52.1 | Improved |
| Add unified representation without zero-shot translation (ablation) | E-FT · CUB | average incremental accuracy: 38.7 -> 9.5 | Hurt |
| Add zero-shot translation (ablation) | E-LwF · CUB | average incremental accuracy: 44.3 -> 52.6 | Improved |
| Add zero-shot translation and unified representation (ablation) | E-LwF · CUB | average incremental accuracy: 44.3 -> 55.0 | Improved |
| Add unified representation without zero-shot translation (ablation) | E-LwF · CUB | average incremental accuracy: 44.3 -> 9.3 | Hurt |
| Add zero-shot translation and unified representation (ablation) | E-EWC · CUB | average incremental accuracy: 51.8 -> 58.1 | Improved |
| Add unified representation without zero-shot translation (ablation) | E-EWC · CUB | average incremental accuracy: 51.8 -> 9.3 | Hurt |
| Add zero-shot translation and unified representation (ablation) | E-MAS · CUB | average incremental accuracy: 48.4 -> 54.8 | Improved |
| Add zero-shot translation (ablation) | E-FT · CIFAR100 | average incremental accuracy: 6.6 -> 7.4 | Improved |
| Add zero-shot translation and unified representation (ablation) | E-FT · CIFAR100 | average incremental accuracy: 6.6 -> 8.5 | Improved |
| Add unified representation without zero-shot translation (ablation) | E-LwF · CIFAR100 | average incremental accuracy: 40.9 -> 6.8 | Hurt |
Closest papers by meaning
- Semantic Drift Compensation for Class-Incremental Learning (CVPR 2020 · similarity 0.90)
- SimZSL: Zero-Shot Learning Beyond a Pre-defined Semantic Embedding Space (IJCV 2025 · similarity 0.89)
- MetaZSCIL: A Meta-Learning Approach for Generalized Zero-Shot Class Incremental Learning (AAAI 2023 · similarity 0.89)
- Class-Incremental Learning: Survey and Performance Evaluation on Image Classification (TPAMI 2022 · similarity 0.89)
- Attentive Region Embedding Network for Zero-Shot Learning (CVPR 2019 · similarity 0.89)
- Tripartite Weight-Space Ensemble for Few-Shot Class-Incremental Learning (CVPR 2025 · similarity 0.87)
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.
