AI Paper Atlas

Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning

Invertible Zero-Shot Recognition Flows

Yuming Shen, Jie Qin, Lei Huang, Li Liu, Fan Zhu, Ling Shao

ECCV 2020 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Generative feature synthesis for zero-shot · 123 citations (OpenAlex)

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 changedWhereMeasuredEffect
add centralizing loss L_C (ablation)IZF-Softmax · AwA1 (GZSL)GZSL harmonic mean H: 68.8 -> 69.6Improved
add centralizing loss L_C (ablation)IZF-NBC · AwA1 (GZSL)GZSL harmonic mean H: 51.4 -> 65.4Improved
add negative-MMD loss L_iMMD (ablation)IZF-Softmax · AwA1 (GZSL)GZSL harmonic mean H: 64.3 -> 69.6Improved
add negative-MMD loss L_iMMD (ablation)IZF-NBC · AwA1 (GZSL)GZSL harmonic mean H: 60.7 -> 65.4Improved
add both L_C and L_iMMD (ablation)IZF-Softmax · AwA1 (GZSL)GZSL harmonic mean H: 51.4 -> 69.6Improved
add both L_C and L_iMMD (ablation)IZF-NBC · AwA1 (GZSL)GZSL harmonic mean H: 52.7 -> 65.4Improved
use flow instead of CVAE (ablation)IZF-NBC · AwA1 (GZSL)GZSL harmonic mean H: 41.8 -> 65.4Improved
use flow instead of CVAE (ablation)IZF-Softmax · AwA1 (GZSL)GZSL harmonic mean H: 48.5 -> 69.6Improved
use negative instead of positive MMD (ablation)IZF-NBC · AwA1 (GZSL)GZSL harmonic mean H: 33.0 -> 65.4Improved
use negative instead of positive MMD (ablation)IZF-Softmax · AwA1 (GZSL)GZSL harmonic mean H: 57.4 -> 69.6Improved
Inverse Multiquadratic kernel instead of Gaussian kernel (ablation)IZF-Softmax · AwA1 (GZSL)GZSL harmonic mean H: 69.5 -> 69.6No clear effect
Inverse Multiquadratic kernel instead of Gaussian kernel (ablation)IZF-NBC · AwA1 (GZSL)GZSL harmonic mean H: 62.9 -> 65.4Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy