The absolute necessity of enhancing optical imagery without expensive hardware is an essential concept in image processing. Multi-exposure image fusion (MEF) ensures an amalgamation of features from varying exposures onto a single fused outcome delivering superior visual traits to the consumer. Recent advances in the area of image processing demand MEF to perform on compromised and limited inputs. However, the pace of handling them with real-time speed and a close-to-humane perception quality has always been undermined. As a solution, the proposed methodology is introduced that overcomes this situation within a reasonable timeframe while delivering prominent qualities. The strategy utilizes a conditional-GAN (cGAN) network to construct a model that advances adaptive exposure correction onto input images. These corrected images are proceeded to refine the actual inputs using an adaptive blending operation. Finally, with the addition of selected weights, a fusion is performed to construct the fused images. The resultant images are compared with the existing MEF techniques and have been verified to deliver superior outcomes on qualitative and quantitative analysis.