Authors: Dongyuan Li, Zhen Wang, Yankai Chen, Renhe Jiang, Weiping Ding, Manabu Okumura

Published on: May 01, 2024

Impact Score: 7.4

Arxiv code: Arxiv:2405.00334

Summary

• What is new: A comprehensive survey on deep learning-based active learning (DAL) that systematically categorizes DAL methods and provides an analysis of their applications in various fields.
• Why this is important: The scarcity of survey papers on DAL, despite its growing popularity and broad applicability.
• What the research proposes: Conducting an advanced and comprehensive survey on DAL, including a formal definition, taxonomy of methods, applications, challenges, and perspectives.
• Results: This survey presents a detailed taxonomy of DAL methods, their strengths and weaknesses, and summarizes applications in NLP, CV, and DM. It also identifies current challenges and offers perspectives for future research.

Technical Details

Technological frameworks used: Deep Learning-based Active Learning

Models used: Various deep model architectures, learning paradigms

Data used: Widely used datasets in NLP, CV, and DM

Potential Impact

This research could impact various sectors including technology companies focused on NLP, CV, and DM, and businesses leveraging AI for efficient data labeling and model training.

Want to implement this idea in a business?

We have generated a startup concept here: OptiLearnAI.