863401402

863401402 / image-provenance

Public

AI image detector and watermark disruption toolkit (client-side, WIP)

17
6
69% credibility
Found May 12, 2026 at 17 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
JavaScript
AI Summary

A client-side web tool for detecting AI-generated images, inspecting metadata including GPS and camera details, performing frequency-domain analysis, and converting images to mimic real camera photos by stripping markers and injecting fake details.

How It Works

1
👀 Find the image checker

You hear about a free online tool that checks if pictures are made by AI or taken with a real camera, all without sending your photos anywhere.

2
🌐 Open the page

Click the demo link in your browser and see a simple screen ready for your picture.

3
📤 Upload your picture

Drag your photo onto the page or pick it from your files, and watch it appear on screen.

4
🔍 Discover the truth

In seconds, you get a clear report: is it AI-generated, what camera took it if real, any location clues, and hidden signs.

5
Choose your next move
📊
Explore more details

Switch tabs to see frequency patterns, full file info, or maps for privacy risks.

🧹
Clean the photo

Pick a camera type, tweak settings, and create a version that looks like a real shot from your phone.

6
💾 Save your work

Download the analysis report or the cleaned photo to your device.

✅ Pictures verified and safe

Now you know your image's story and have a private, cleaned copy ready to share without worries.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 17 to 17 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is image-provenance?

This JavaScript toolkit is a client-side image provenance analysis tool for detecting AI-generated images and disrupting watermarks. Upload any photo to get multi-layer checks—C2PA metadata, EXIF hints from tools like Midjourney or Google SynthID, pixel-level frequency analysis with 65 features, and detailed breakdowns across tabs for provenance, freq domain, metadata, and conversion. It solves the black-box problem of server-side image detectors by running 100% in-browser, keeping your image detector ai free and private.

Why is it gaining traction?

Unlike basic image detector online tools that just spit out a score, this github image repo delivers image provenance analysis at scale with viridis heatmaps, radial spectra, and evidence chains—letting you inspect why it flags fakes. The killer hook is the conversion pipeline: strip C2PA/watermarks, re-encode, and inject realistic EXIF from 17 camera profiles (iPhone 17 Pro to Nikon Z8), turning AI outputs into "real photo" mimics without visible changes. Zero-build setup via local HTTP server or online demo makes it dead simple for quick tests.

Who should use this?

Journalists fact-checking image detector fake claims before publishing; frontend devs adding client-side image provenance checks to upload forms; security researchers probing AI detector google or macro weaknesses; privacy hawks scrubbing GPS/EXIF from personal shots.

Verdict

Grab it for browser-based image detector ai experiments—strong docs, demo, and UI punch above 17 stars, though WIP status and 0.699999988079071% credibility score mean expect rough edges like mobile perf tweaks. Solid for eval, not prod yet.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.