Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Transductive Zero-Shot Action Recognition by Word-Vector Embedding
Xun Xu, Timothy M. Hospedales, Shaogang Gong
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 data augmentation from auxiliary datasets (ablation) | RR + NN · HMDB51 | accuracy (%): 14.5 -> 20.4 | Improved |
| add data augmentation from auxiliary datasets (ablation) | RR + NN · UCF101 | accuracy (%): 11.7 -> 15.7 | Improved |
| add data augmentation from auxiliary datasets (ablation) | RR + NN · CCV | mAP (%): 20.7 -> 30.3 | Improved |
| add data augmentation from auxiliary datasets (ablation) | RR + NN · USAA | accuracy (%): 29.5 -> 28.2 | Hurt |
| replace ridge regression with manifold regularized regression (with self-training) (ablation) | NN matching · HMDB51 | accuracy (%): 17.0 -> 18.6 | Improved |
| replace ridge regression with manifold regularized regression (with self-training) (ablation) | NN matching · UCF101 | accuracy (%): 15.9 -> 17.6 | Improved |
| add self-training of class prototypes (ablation) | RR + NN · UCF101 | accuracy (%): 11.7 -> 15.9 | Improved |
| add normalised nearest neighbour hubness correction (ablation) | RR · UCF101 | accuracy (%): 11.7 -> 13.9 | Improved |
| use NRM or GC instead of NN matching (ablation) | mAP datasets · Olympic Sports and CCV | mAP (%) | No clear effect |
| add self-training and manifold regularization (ablation) | RR + NN · TRECVID MED 2013 | mAP (%): 1.18 -> 1.38 | Improved |
| add self-training to SJE (ablation) | SJE · HMDB51 | accuracy (%): 12.0 -> 10.5 | Hurt |
| add self-training to ConSE (ablation) | ConSE · HMDB51 | accuracy (%): 15.0 -> 15.4 | Improved |
Closest papers by meaning
- Multi-Task Zero-Shot Action Recognition with Prioritised Data Augmentation (ECCV 2016 · similarity 0.94)
- Cross-Modal Representation Learning for Zero-Shot Action Recognition (CVPR 2022 · similarity 0.91)
- Reformulating Zero-shot Action Recognition for Multi-label Actions (NeurIPS 2021 · similarity 0.91)
- ZeroDiff: Solidified Visual-semantic Correlation in Zero-Shot Learning (ICLR 2025 · similarity 0.89)
- ZeroMamba: Exploring Visual State Space Model for Zero-Shot Learning (AAAI 2025 · similarity 0.88)
- Part-aware Unified Representation of Language and Skeleton for Zero-shot Action Recognition (CVPR 2024 · similarity 0.88)
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.
