Humans are remarkably fast and accurate at detecting faces, even across diverse scene contexts. While scene context is known to influence face detection, the neural mechanisms underlying this effect remain unclear. Here, we investigated how scene previews modulate neural responses during face detection using EEG. A set of 784 scene images containing a single face (left or right) served as target stimuli, with corresponding faceless versions as previews. In an EEG experiment, participants (N=28) saw either a scene preview (preview condition) or a neutral gray screen (no-preview condition) for 250 ms, followed by the target scene. Participants identified whether the scene contained a face (20% foils). Temporal multivariate decoding of neural activity (-100 to 1100 ms) using SVMs revealed higher decoding accuracy of face position (left vs. right) in the no-preview condition between 400–600 ms, suggesting enhanced neural processing of spatial face information without prior scene context. In contrast, time-frequency decoding showed greater accuracy in the preview condition in the alpha (8-13 Hz) band—often associated with top-down processing—highlighting the role of alpha in encoding face position under the influence of scene context. These findings uncover distinct neural mechanisms by which prior scene information influences face detection, with ERP decoding reflecting bottom-up integration processes and alpha activity capturing context-driven attentional enhancement. Our study advances understanding of face perception in naturalistic environments, emphasizing the interplay between scene context and neural dynamics.