On the limits of Conditional Generative Adversarial Neural Networks to reconstruct the identification of inhabitants from IoT low-resolution thermal sensors

作者：

Highlights：

• Low resolution thermal face images into color visible-spectrum images.

• Evaluating privacy in a smart home from GANs with thermal sensors.

• Using Pix2Pix CGANs with residual networks and face recognition modules.

• Design and implementation of a MultiGPU CGAN system.

摘要

•Low resolution thermal face images into color visible-spectrum images.•Evaluating privacy in a smart home from GANs with thermal sensors.•Using Pix2Pix CGANs with residual networks and face recognition modules.•Design and implementation of a MultiGPU CGAN system.

论文关键词：Generative Adversarial Networks (GANs),Image translation,Thermal image,Face recognition,Privacy

论文评审过程：Received 2 July 2021, Revised 4 March 2022, Accepted 25 April 2022, Available online 6 May 2022, Version of Record 14 May 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117356