Progressive polarization based reflection removal via realistic training data generation
作者：
Highlights：
• A realistic and diversified training dataset (POL) is constructed by optical modeling for reflection obstructed image.
• A progressive polarization based reflection removal network (P2R2Net) with up-to-date neural network units is proposed .
• The improvement of the proposed method for real-world images is verified by comparisons with the state-of-the-art methods.
• Robustness analysis on reflection-tainted synthesis images suggests our method possesses excellent generalization ability.
摘要
•A realistic and diversified training dataset (POL) is constructed by optical modeling for reflection obstructed image.•A progressive polarization based reflection removal network (P2R2Net) with up-to-date neural network units is proposed .•The improvement of the proposed method for real-world images is verified by comparisons with the state-of-the-art methods.•Robustness analysis on reflection-tainted synthesis images suggests our method possesses excellent generalization ability.
论文关键词：Deep learning,Reflection removal,Polarization,Progressive network,Convolutional neural networks
论文评审过程：Received 20 February 2021, Revised 8 July 2021, Accepted 9 December 2021, Available online 11 December 2021, Version of Record 20 December 2021.
论文官网地址：https://doi.org/10.1016/j.patcog.2021.108497
