WRGPruner: A new model pruning solution for tiny salient object detection

作者：

Highlights：

• Define the new concept salient energy level (SEL) to assess the parameters' ability to distinguish importance.

• Propose the novel method WRGPruner that simultaneously consider the synthesis of three dimensions.

• Mathematically prove the effectiveness of the WRGPruner for tiny salient object detection.

• Construct a tiny salient object dataset (TSOD) and its extension version TSOD-S, TSOD-M, and TSOD-L.

摘要

•Define the new concept salient energy level (SEL) to assess the parameters' ability to distinguish importance.•Propose the novel method WRGPruner that simultaneously consider the synthesis of three dimensions.•Mathematically prove the effectiveness of the WRGPruner for tiny salient object detection.•Construct a tiny salient object dataset (TSOD) and its extension version TSOD-S, TSOD-M, and TSOD-L.

论文关键词：Model compression,Model pruning,Salient objects detection,Small objects detection,Computer vision

论文评审过程：Received 4 January 2021, Revised 15 February 2021, Accepted 16 February 2021, Available online 20 February 2021, Version of Record 4 March 2021.

论文官网地址：https://doi.org/10.1016/j.imavis.2021.104143