ScholarMate
English简体中文繁體中文
客服熱線:400-1616-289
登入註冊
期刊論文

Active Fine-Tuning From gMAD Examples Improves Blind Image Quality Assessment

Wang, Zhihua*; Ma, Kede
IEEE Transactions on Pattern Analysis and Machine Intelligence2022-9-1DOI: 10.1109/TPAMI.2021.3071759
9被引5閱讀
引用收藏
分享

摘要

The research in image quality assessment (IQA) has a long history, and significant progress has been made by leveraging recent advances in deep neural networks (DNNs). Despite high correlation numbers on existing IQA datasets, DNN-based models may be easily falsified in the group maximum differentiation (gMAD) competition. Here we show that gMAD examples can be used to improve blind IQA (BIQA) methods. Specifically, we first pre-train a DNN-based BIQA model using multiple noisy annotators, and fine-tune it on multiple synthetically distorted images, resulting in a "top-performing" baseline model. We then seek pairs of images by comparing the baseline model with a set of full-reference IQA methods in gMAD. The spotted gMAD examples are most likely to reveal the weaknesses of the baseline, and suggest potential ways for refinement. We query human quality annotations for the selected images in a well-controlled laboratory environment, and further fine-tune the baseline on the combination of human-rated images from gMAD and existing databases. This process may be iterated, enabling active fine-tuning from gMAD examples for BIQA. We demonstrate the feasibility of our active learning scheme on a large-scale unlabeled image set, and show that the fine-tuned quality model achieves improved generalizability in gMAD, without destroying performance on previously seen databases.

學科與關鍵詞

關鍵詞
Computational modelingDatabasesAdaptation modelsTrainingPredictive modelsTask analysisImage qualityBlind image quality assessmentdeep neural networksgMAD competitionactive learning
相關
載入中…

文獻載入失敗,請重試

重試

發表資訊

論文狀態
公開發表
期刊
IEEE Transactions on Pattern Analysis and Machine Intelligence
發表日期
2022-09-01
卷期
第 44 卷, 第 9 期
頁碼
4577-4590
DOI
10.1109/TPAMI.2021.3071759
語言
ENGLISH
期刊等級
1区
成果收錄
SCIE
資助標註
National Natural Science Foundation of China [62071407]; CityU Start-up Grant [7200630]

作者

  • W

    Wang, Zhihua (通訊) (第一作者)

  • M

    Ma, Kede

登入解鎖更多功能

  • 收藏這篇,隨時在「我的 · 收藏」找回
  • 認領你的成果,生成推廣素材
  • 關注作者獲取最新成果
  • 加入項目組、協同寫作
註冊 / 登入

產品服務

  • 科研之友
  • 創新城
  • 科創雲

服務支援

  • 幫助中心
  • 隱私政策
  • 服務條款

聯絡方式

線上客服:【】
客服熱線:400-1616-289
電子郵件:support@scholarmate.com

關注或下載科研之友

微信 QR Code
微信公眾號
客戶端下載 QR Code
下載客戶端
科研成果科研人員科研機構科研動態愛瑞思軟件

©2026 深圳市科研之友网络服务有限公司

公安備案圖示粤公网安备 44030502000213
粤ICP备 16046710 号粤B2-20110417