All Projects → roboflow-ai → neuralhash-collisions

roboflow-ai / neuralhash-collisions

Licence: MIT license
A catalog of naturally occurring images whose Apple NeuralHash is identical.

Programming Languages

javascript
184084 projects - #8 most used programming language
python
139335 projects - #7 most used programming language

Projects that are alternatives of or similar to neuralhash-collisions

AppleNeuralHash2ONNX
Convert Apple NeuralHash model for CSAM Detection to ONNX.
Stars: ✭ 1,457 (+904.83%)
Mutual labels:  apple, neuralhash
crossbow
Cross-Platform Rust Toolkit for Games 🏹
Stars: ✭ 80 (-44.83%)
Mutual labels:  apple
react-native-single-select
Customizable & Easy to Use Single Select Library for React Native
Stars: ✭ 74 (-48.97%)
Mutual labels:  apple
dotAPNS
dotAPNS is a library used to send push notifications to Apple devices using Apple Push Notification service via HTTP/2 API.
Stars: ✭ 80 (-44.83%)
Mutual labels:  apple
flying-apple
Just to keep track of nice content and new announcements related to Apple products and Swift
Stars: ✭ 45 (-68.97%)
Mutual labels:  apple
LPThumbnailView
A thumbnail view for iOS to give context to multiple images/videos using thumbnails and counter.
Stars: ✭ 54 (-62.76%)
Mutual labels:  apple
mbpfan
A simple daemon to control fan speed on all MacBook/MacBook Pros (probably all Apple computers) for Linux Kernel 3 and newer
Stars: ✭ 680 (+368.97%)
Mutual labels:  apple
BDLocalizedDevicesModels
Apple devices model names localized.
Stars: ✭ 23 (-84.14%)
Mutual labels:  apple
brewfile
🍎 Brewfile to install softwares in macOS for engineers
Stars: ✭ 37 (-74.48%)
Mutual labels:  apple
MSession
A simple and sophisticated session and authentication solution written in Swift
Stars: ✭ 26 (-82.07%)
Mutual labels:  apple
AppleScripts
💬 A collection of useful AppleScripts I've discovered on the web and written.
Stars: ✭ 17 (-88.28%)
Mutual labels:  apple
apns2
Node client for connecting to Apple's Push Notification Service using the new HTTP/2 protocol with JSON web tokens
Stars: ✭ 66 (-54.48%)
Mutual labels:  apple
scala-phash
Image comparison by hash codes
Stars: ✭ 16 (-88.97%)
Mutual labels:  perceptual-hashing
apple-knowledge
A collection of reverse engineered Apple things, as well as a machine-readable database of Apple hardware
Stars: ✭ 338 (+133.1%)
Mutual labels:  apple
onepile
Playground for the future of private notes and document management
Stars: ✭ 41 (-71.72%)
Mutual labels:  apple
Clover GA-H97-D3H
No longer maintained, now OpenCore is supported
Stars: ✭ 25 (-82.76%)
Mutual labels:  apple
mac os scripts
Some scripts for automating Windows domain stuff for MacOS machines
Stars: ✭ 27 (-81.38%)
Mutual labels:  apple
albumArtCreator
A simple tool to make it easy to create good looking album art
Stars: ✭ 26 (-82.07%)
Mutual labels:  apple
tl-create
tl-create is a cross-platform command line tool to create a X.509 trust list from various trust stores. (Keywords: CABFORUM, eIDAS, WebPKI)
Stars: ✭ 32 (-77.93%)
Mutual labels:  apple
SwiftUIFormValidator
Declarative form validator for SwiftUI.
Stars: ✭ 34 (-76.55%)
Mutual labels:  apple

neuralhash-collisions

A catalog of naturally occurring images whose Apple NeuralHash is identical.

See this NeuralHash collision blog post with full details and background.

Contributing

PRs are welcome and encouraged! You can find the code to run NeuralHash on images in this repo. Please do not submit artificially generated adversarial examples. This repo is meant to document natural collisions found in the wild.

Please follow the example format (add a folder in collisions named with the colliding hash and containing the example images with a README cataloging information about their provenance). Ideally provide evidence that these images existed prior to the public divulgence of the NeuralHash weights to prove that they are not artificial.

Suggested Methodology

There are many open source image datasets which contain a wide variety of images.

If you generate a NeuralHash for each one, store the image path / hash pairs in a text file, sort them by hash, and compare adjacent lines' hashes, you can quickly check n^2 image pairs (where n is the size of the dataset).

There are some example scripts you may adapt in the utils directory.

Hash Lists

You may also compare your hashes against the ImageNet hashes using the txt file archived in the imagenet directory.

If you check a large, public corpus please PR the hash list so others can easily compare them for collisions across sources.

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].