Densely connected convolutional network block based autoencoder for panorama map compression

作者：

Highlights：

• Proposed a dense block based autoencoder.

• Designed a cubic projection based block partition scheme.

• Designed a weighted loss function.

• Proposed a greedy block-wise training method.

摘要

•Proposed a dense block based autoencoder.•Designed a cubic projection based block partition scheme.•Designed a weighted loss function.•Proposed a greedy block-wise training method.

论文关键词：Autoencoder,Dense block,Neural network,Panorama map

论文评审过程：Received 14 November 2018, Revised 23 October 2019, Accepted 24 October 2019, Available online 28 October 2019, Version of Record 30 October 2019.

论文官网地址：https://doi.org/10.1016/j.image.2019.115678