There Is One Key Hidden in This Room. Can You Find It?

The scene below is composed from data, the same way the spot-the-difference scenes are. A list of objects is placed with rejection sampling so that nothing overlaps anything else, and then one target object — of a kind that appears nowhere else in the picture — is dropped into a free space.

Its coordinates are recorded. That is the answer.

Three rules this puzzle follows

The target is never cut off by the edge. A validator checks that the target sits at least a third of its own width inside the picture on every side. Hiding something half off the canvas is not a puzzle, it is a trick.

The target appears exactly once. The generator is told to exclude that object kind from the rest of the scene, and the validator counts every object of that kind before publication. If it finds two, the puzzle fails.

The alt text gives nothing away. This is the one most often missed. A screen reader user should get the same puzzle as everyone else, which means the description of the challenge image must not say where the target is — or even name it. So the validator rejects any challenge alt text containing the target’s name or any positional word: left, right, top, bottom, upper, lower, corner, centre, behind.

The solution image is different. Its alt text describes exactly where the target is, because by then the puzzle is over.

If you are stuck

Do not scan for the shape. Scan for the colour, which is usually the faster route in a scene with this much going on, then check each candidate.

The reveal circles it and tells you roughly where it was.