ARRPNGAN: Text-to-image GAN with attention regularization and region proposal networks

作者：

Highlights：

• Text-to-image synthesis combines textual and image features.

• Generative adversarial network shows promising results in image synthesis.

• Attention mechanism is effective at locating target objects and features.

• Region proposal network can eliminate background interferences.

• Proposing networks for synthesizing real-world images.

摘要

•Text-to-image synthesis combines textual and image features.•Generative adversarial network shows promising results in image synthesis.•Attention mechanism is effective at locating target objects and features.•Region proposal network can eliminate background interferences.•Proposing networks for synthesizing real-world images.

论文关键词：Text-to-image synthesis,Generative adversarial network,Attention model,Region proposal network

论文评审过程：Received 26 October 2021, Revised 15 January 2022, Accepted 18 April 2022, Available online 30 April 2022, Version of Record 11 May 2022.

论文官网地址：https://doi.org/10.1016/j.image.2022.116728