The critical heat flux (CHF) is an essential safety boundary in boiling heat transfer processes employed in high heat flux thermal-hydraulic systems. Identifying CHF is vital for preventing equipment damage and ensuring overall system safety, yet it is challenging due to the complexity of the phenomena. For an in-depth understanding of the complicated phenomena, various methodologies have been devised, but the acquisition of high-resolution data is limited by the substantial resource consumption required. This study presents a data-driven, image-to-image translation method for reconstructing thermal data of a boiling system at CHF using conditional generative adversarial networks (cGANs). The supervised learning process relies on paired images, which include total reflection visualizations and infrared thermometry measurements obtained from flow boiling experiments. Our proposed approach has the potential to not only provide evidence connecting phase interface dynamics with thermal distribution but also to simplify the laborious and time-consuming experimental setup and data-reduction procedures associated with infrared thermal imaging, thereby providing an effective solution for CHF diagnosis.