White light phase shifting interference (WL-PSI) microscopy is a prominent technique for high-resolution quantitative phase imaging (QPI) of industrial and biological specimens. Highly sensitive and accurate phase measurement is possible using WL-PSI because of low coherence properties of light source and the phase shifting algorithm. Multiple phase-shifted interferograms with accurate phase shift is obligatory in WL-PSI for measuring accurate phase map of the object. However, phase error occurs during the experimentation due to environmental perturbation and recording multiple frames. Here, we present a single-shot phase shifting interferometric technique for accurate phase measurement using filtered WL-PSI and deep neural network (DNN). The method is implemented by training the DNN to generate the phase shifted frames from a single recorded interferogram that are equivalent to experimentally recorded phase shifted interferograms. We simulate and experimentally demonstrate the robustness of the proposed framework on strip step-like waveguide structure. The results show precise matching of reconstructed phase map from the DNN generated phase shifted interferograms and experimentally recorded interferograms. The current WLPSI+DNN approach may further strengthen QPI techniques by high resolution phase recovery using single frame for different biomedical applications