Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Generalized Zero-Shot Learning via Over-Complete Distribution
Rohit Keshari, Richa Singh, Mayank Vatsa
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 OCD to OBTL training (ablation) | OBTL classifier · AWA2, SUN, and CUB PS | classification accuracy (%): 65.8 -> 70.9; 56.4 -> 62; 54.5 -> 60.5 | Improved |
| Add OCD to CL training (ablation) | CL classifier · AWA2, SUN, and CUB PS | classification accuracy (%): 65.3 -> 66.5; 56.2 -> 57.6; 53.7 -> 56.8 | Improved |
| Add CL to OCD+OBTL (ablation) | OCD-CVAE · AWA2, SUN, and CUB PS | classification accuracy (%): 70.9 -> 71.3; 62 -> 62.1; 60.5 -> 60.9 | Improved |
| Add OBTL to OCD+CL (ablation) | OCD-CVAE · AWA2, SUN, and CUB PS | classification accuracy (%): 66.5 -> 71.3; 57.6 -> 62.1; 56.8 -> 60.9 | Improved |
| Use only OBTL or only CL, without OCD (ablation) | OCD-CVAE · AWA2, SUN, and CUB PS | classification accuracy (%): 65.8, 56.4, 54.5; 65.3, 56.2, 53.7 | Hurt |
| Generate OCD and train with the proposed framework | Zero-shot classifier · AWA2 PS | classification accuracy (%): 69.5 -> 71.3 | Improved |
| Generate OCD and train with the proposed framework | Zero-shot classifier · CUB PS | classification accuracy (%): 59.6 -> 60.3 | Improved |
| Generate OCD and train with the proposed framework | Zero-shot classifier · SUN PS | classification accuracy (%): 63.4 -> 63.5 | No clear effect |
| Use the proposed OCD-CVAE framework | GZSL classifier · AWA2 GZSL | harmonic mean H (%): 63.9 -> 65.7 | Improved |
| Use the proposed OCD-CVAE framework | GZSL classifier · CUB GZSL | harmonic mean H (%): 52.4 -> 51.3 | Hurt |
| Use the proposed OCD-CVAE framework | GZSL classifier · SUN GZSL | harmonic mean H (%): 53.4 -> 43.8 | Hurt |
| Generate 400 to 600 OCD samples for training | AWA2 regressor · AWA2 PS | classification accuracy | Improved |
Closest papers by meaning
- Generalized Zero-Shot Learning via Synthesized Examples (CVPR 2018 · similarity 0.91)
- Zero-Shot Learning via Class-Conditioned Deep Generative Models (AAAI 2018 · similarity 0.90)
- A Variational Autoencoder with Deep Embedding Model for Generalized Zero-Shot Learning (AAAI 2020 · similarity 0.89)
- Data Distribution Distilled Generative Model for Generalized Zero-Shot Recognition (AAAI 2024 · similarity 0.89)
- Data-Free Generalized Zero-Shot Learning (AAAI 2024 · similarity 0.87)
- Improving Generalized Zero-Shot Learning by Exploring the Diverse Semantics from External Class Names (CVPR 2024 · 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.
