The low-frequency component of the upcoming Square Kilometre Array Observatory (SKA-Low) will be sensitive enough to construct 3D tomographic images of the 21-cm signal distribution during reionization. However, foreground contamination poses challenges for detecting this signal, and image recovery will heavily rely on effective mitigation methods. We introduce SERENEt, a deep-learning framework designed to recover the 21-cm signal from SKA-Low’s foreground-contaminated observations, enabling the detection of ionized (Hii) and neutral (Hi) regions during reionization. SERENEt can recover the signal distribution with an average accuracy of 75percnt at the early stages ($\overline{x}\mathrm{HI}\simeq 0.9$) and up to 90percnt at the late stages of reionization ($\overline{x}\mathrm{HI}\simeq 0.1$). Conversely, Hi region detection starts at 92percnt accuracy, decreasing to 73percnt as reionization progresses. Beyond improving image recovery, SERENEt provides cylindrical power spectra with an average accuracy exceeding 93percnt throughout the reionization period. We tested SERENEt on a 10-degree field-of-view simulation, consistently achieving better and more stable results when prior maps were provided. Notably, including prior information about Hii region locations improved 21-cm signal recovery by approximately 10percnt. This capability was demonstrated by supplying SERENEt with ionizing source distribution measurements, showing that high-redshift galaxy surveys of similar observation fields can optimize foreground mitigation and enhance 21-cm image construction.