Recognition: unknown
Generative Adversarial Active Learning
read the original abstract
We propose a new active learning by query synthesis approach using Generative Adversarial Networks (GAN). Different from regular active learning, the resulting algorithm adaptively synthesizes training instances for querying to increase learning speed. We generate queries according to the uncertainty principle, but our idea can work with other active learning principles. We report results from various numerical experiments to demonstrate the effectiveness the proposed approach. In some settings, the proposed algorithm outperforms traditional pool-based approaches. To the best our knowledge, this is the first active learning work using GAN.
This paper has not been read by Pith yet.
Forward citations
Cited by 3 Pith papers
-
Synthetic Designed Experiments for Diagnosing Vision Model Failure
SDRS uses designed experiments and ANOVA decomposition on synthetic data to identify Type I coverage gaps and Type II spurious dependencies in vision models, then generates targeted data to improve performance.
-
An Analysis of Active Learning Algorithms using Real-World Crowd-sourced Text Annotations
Real crowd-sourced text annotations are used to test eight active learning techniques, showing their behavior under actual noisy labeling and refusals unlike simulated oracles.
-
Portable Active Learning for Object Detection
PAL is a portable active learning method for object detection that uses class-specific logistic classifiers for uncertainty and image-level diversity to select annotation batches, showing better label efficiency than ...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.