Item logo image for Obscur

Obscur

ExtensionTools
Item media 4 (screenshot) for Obscur
Item media 1 (screenshot) for Obscur
Item media 2 (screenshot) for Obscur
Item media 3 (screenshot) for Obscur
Item media 4 (screenshot) for Obscur
Item media 1 (screenshot) for Obscur
Item media 1 (screenshot) for Obscur
Item media 2 (screenshot) for Obscur
Item media 3 (screenshot) for Obscur
Item media 4 (screenshot) for Obscur

Overview

Detect AI images and video frames patch by patch — the obscur.rao.nyc detector, one right-click or tap away. A research preview.

Obscur — verify reality, one image at a time. AI-generated images are everywhere, and the tells that used to give them away — extra fingers, melted text — are disappearing fast. Obscur reads an image the way a forensic tool would: not "does this look fake," but whether the physical fingerprints a real camera sensor leaves behind are actually there. What it does Right-click any image on the web and choose Obscur. In a couple of seconds you get a verdict — real, AI-generated, or a synthetic render — laid over the image itself, patch by patch, so you can see where the evidence is rather than taking a number on faith. Hover over a large image or a video and a small Obscur medallion appears — one tap runs the same analysis. On video, Obscur grabs the current frame straight from the player, so you can check a clip mid-scroll without saving anything. What you see A real ↔ AI scale with both the weighted verdict and the tile-by-tile majority vote, so you can tell a confident call from a close one. Heatmap overlays — AI likelihood per region, the camera sensor map, and which patches actually drove the verdict. The specialist stack — the individual models that voted (sensor noiseprint, camera warp-fusion, AI-art detectors, render arbiter) and what each one concluded. When they disagree, you see it. Capture signature — whether a genuine camera-capture fingerprint is present at all. Why install it Because "trust me" isn't evidence. Obscur shows its work: every verdict comes with the regions, the scores, and the dissenting specialists behind it. It's the same detector that runs at obscur.rao.nyc/detect, one right-click away on any page — no uploading, no tab-switching, no account. Honest limits — please read This is a research preview. The detector is tuned to the physics of the iPhone camera sensor, and it's strongest on ordinary photographs. Treat verdicts on screenshots, scans, heavily edited images, and non-photographic art as advisory, not conclusive. It's a lens, not a lie detector — each release iterates it toward a general reality verifier. Privacy Images you choose to analyze are sent to obscur.rao.nyc for inference and are not stored. Obscur only ever sees the specific image or frame you select — it never scans pages, reads your browsing, or acts on its own. Settings The hover medallion on images can be turned off entirely, or given a minimum size so avatars and thumbnails stay quiet while full-size photos stay one tap away.

Details

  • Version
    0.1.3
  • Updated
    July 19, 2026
  • Offered by
    Rao Studios
  • Size
    87.1KiB
  • Languages
    English (United States)
  • Developer
    Email
    ritesh@rao.nyc
  • Non-trader
    This developer has not identified itself as a trader. For consumers in the European Union, please note that consumer rights do not apply to contracts between you and this developer.

Privacy

Manage extensions and learn how they're being used in your organization
The developer has disclosed that it will not collect or use your data. To learn more, see the developer’s privacy policy.

This developer declares that your data is

  • Not being sold to third parties, outside of the approved use cases
  • Not being used or transferred for purposes that are unrelated to the item's core functionality
  • Not being used or transferred to determine creditworthiness or for lending purposes

Support

For help with questions, suggestions, or problems, visit the developer's support site

Google apps