Once upon a time, a computer could tell you virtually nothing about an image beyond its file format, size, and color palette. These days, powerful image recognition systems are a part of our everyday lives. They See Your Photos is a simple website that shows you just how much these systems can interpret from a regular photo.

The website simply takes your image submission, runs it through the Google Vision API, and spits back out a description of the image. I tried it out with a photograph of myself, and was pretty impressed with what the vision model saw:

The photo is taken in a lush green forest, with tall trees dominating the background. The foreground features a person, who appears to be the subject of the photograph. The lighting suggests it might be daytime, and the overall color palette is heavily saturated with shades of green, almost artificial in appearance. There’s also some dried vegetation visible to the left, suggesting a natural setting that is possibly a park or woodland area.

The subject is a young to middle-aged Caucasian male with shoulder-length, light-colored hair. He seems serious, perhaps pensive or slightly uneasy. His clothing —a green and yellow checkered shirt over a green and black striped shirt—suggests a casual or outdoorsy lifestyle. He might be of middle to lower-middle class economic standing. It looks like he’s crouching slightly, possibly for the picture. The image lacks metadata on the camera device used or the time the photo was taken. He appears to be alone in the photo, indicating an individualistic or solitary experience.

The saturation level of the greens and yellows is unusually high, hinting at possible digital editing post-capture. There is a very slight blur, particularly noticeable in the background which could be from a smaller aperture or shallow depth of field when captured, creating a focus on the subject. The color alteration and seemingly intentional focus on the subject suggest it may not be a candid shot but rather a posed photograph, possibly with an artistic or stylistic goal.

The model did very well—easily determining both the vague type of locale , and the fact that my shirt implies I don’t have a Ferrari at home in my garage. It also picks up on the fact that it was a posed photograph with certain artistic intent.

Back in 2014, the webcomic XKCD stated that it would be an inordinately difficult task for a computer to determine if a digital photo contained a bird. These days, a computer model can tell us what’s in a photo down to the intimate details, and even make amusing assertions as to the lives of the subjects in the image and their intentions. We’ve come a long way, to be sure.

Machine vision is still far from perfect—there are ways to fool systems and areas in which they still don’t stack up to real humans. The only thing we know for certain is that these systems will continue to improve.