Original title: Privacy Leakage from a Thousand Words: Millipixel Location Recovery from Dot Maps
Yuntao Du, Tanishq Pauskar, Hao Wang, Jing Su +1 · September 2026
Maps often show data by placing dots over a geographic area. People use these dot maps to share sensitive spatial trends, but they might reveal too much private information. Researchers created an automated framework called AutoLocate to test this danger. When computers draw smooth dot edges on a screen, they use a process called anti-aliasing. This smoothing process unintentionally hides tiny clues about exact dot coordinates in the pixels. AutoLocate treats location recovery as a black-box puzzle. It repeatedly guesses dot coordinates, draws new test maps, and adjusts its guesses to match the target map's smoothed pixel edges. It works well across varied map styles and datasets. On small-scale United States maps, it recovers coordinates within an average of 1 meter.
Imagine you are trying to find the exact spot where a friend dropped a coin on a sandy beach. The tide has smoothed the sand, but there are tiny, barely visible patterns left by the coin's impact. By repeatedly smoothing sand in different spots and comparing to the original pattern, you can pinpoint the coin's location. In this paper, the dot map is the smoothed sand, and the anti-aliasing artifacts are the tiny patterns; AutoLocate iteratively adjusts guessed dot positions until the rendered map matches the target map's artifacts.
Anyone publishing a dot map of sensitive data—like disease cases or voting patterns—could be unknowingly exposing exact locations. This work shows that even a simple map image leaks sub-meter precision coordinates through rendering artifacts. It also provides tools to assess and reduce that risk before publication.
The abstract does not state explicit limitations, but the attack relies on anti-aliasing artifacts; maps rendered without anti-aliasing may be immune. The experiments cover a range of configurations but not all possible scenarios. The proposed mitigation strategies are not fully evaluated in the abstract.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Privacy Leakage from a Thousand Words: Millipixel Location Recovery from Dot Maps — Yuntao Du, Tanishq Pauskar, Hao Wang et al.
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.