This paper presents a novel machine learning approach to PM image generation of ECD panel for HDR imaging. The high transmittance ECD panel can be employed in front of camera lens to obtain HDR images by providing low and high transmittance over the bright and dark areas, respectively. Additionally, to achieve low power consumption, fast refresh rate, and low fabrication cost, the whole pixels of an ECD panel should be updated simultaneously, which is named as simultaneous PM driving (SPMD) method. Because there is the limitation on the images that are presented by SPMD, it needs to find out the closest PM image to the reference image. The proposed PMNet based on two fully-connected layers learns an end-to-end mapping between the reference images and the PM images. The proposed PMNet is evaluated in MSE and SSIM.