1: \begin{abstract}

2: Brightness order reversal could happen among over-exposed regions of a bright image and under-exposed regions of a dark image if two large-exposure-ratio images are fused directly by using existing multi-scale exposure fusion (MEF) algorithms. This problem can be addressed efficiently by interpolating a virtual image with a medium exposure time. In this paper, a new exposure interpolation algorithm is introduced by combining model driven and data driven exposure interpolation methods. The key idea is to obtain an initial medium-exposure image by using intensity mapping functions (IMFs), while the modeling error is compensated by the data-driven method. Experimental results indicate that the data-driven method is benefited from the model-driven method for fast convergence speed and demand of large training samples. The final interpolated medium-exposure image is significantly improved by employing the hybrid methods in terms of PSNR and SSIM metrics.

3: \end{abstract}

4: