The face images above slowly change between sharp and blurred. When they are sharp, the face on the left looks angry and male, the face on the right female and kind (no generic character attribution implied☺). With increasing blur the expressions change places.
You can manually set the blur level (use the buttons ›sharp‹ or ›blurred‹, or ›stop‹ and drag the slider.
Instead of using the computer blur, you can simply select ›sharp‹ and step away from the screen (or squint, or remove glasses, or use any method that blurs the image).
Using a technique they described 1994, Philippe G. Schyns & Aude Oliva (1999) hid 2 faces in one hybrid image: one image was high-pass filtered (≥ 8 cpd), the other low-pass filtered (≤ 2 cpd) [reproduced with kind permission]. With normal vision, the higher spatial frequency image ‘wins’. When blurring the image (by whatever means), only the low-pass image remains. This method can be applied to many image types, but with a face exchange it seems especially amazing.
Schyns PG, Oliva A (1994) From blobs to boundary edges: Evidence for time and spatial scale dependent scene recognition. Psychol Sci 5:195–200
Schyns PG, Oliva A (1999) Dr. Angry and Mr. Smile: when categorization flexibly modifies the perception of faces in rapid visual presentations. Cognition 69:243–265
Computational Visual Cognition Laboratory Gallery
Thanks to Neal Miller, from whom I learned about this phenomenon.