Welcome to the ClassifierAI Repository!
Description
A Google Chrome Extension that integrates Machine Learning to determine if the image and the dev.to article, that the user is viewing, is AI-Generated. This uses Tensorflow.js to perform Image Classification and Text Classification.
The model is trained using Teachable Machine by Google where it is train from 866 images, that consists of AI and Non-AI images, using the following settings:
- Epochs: 30
- Batch Size: 16
- Learning Rate: 0.0001
Technologies Used
Core Tech: JavaScript HTML5 CSS3 Tailwind CSS
Content Analysis: TensorFlow.js Chart.js
Extension Tooling: Vite CRXJS Docker
How to Run
-
Clone or download the repository and navigate to the project directory.
-
Install the project dependencies:
npm install
-
Build the extension:
npm run build
-
Open Google Chrome and navigate to:
chrome://extensions -
Enable Developer Mode in the top-right corner.
-
Click Load unpacked and select the generated
dist/folder. -
The extension is now installed and ready to use.
Demo
This is the overall demo of the use...
Top comments (38)
Heyyy! Hope you're doing better now Francis!
Thank you for the mention and keep stacking those awesome monthly wins🖖 You're doing great, keep moving forward!!
Stanlee
Thanks Aryan and thanks for checking up on me! Feeling a lot better now :D
That's honestly great to hear.
Don't worry. I am not going anywhere if you are concern :D
Four legends in one frame 🔥
Pretty solid month. Glad the mood is better now! I'd still say a good idea would be for you to start a side-project, just something for yourself, even if it's a meme selector that preps it for you to post, or place in your post? Think that would save you hours in a month 😂
Thanks! Not a bad idea, though I want to focus on speaking well in interviews more. We shall see but thanks for the suggestion :D
Always a good thing to focus on. Being in-charge in an interview is a worthwhile skill, because it makes it clear they need YOU, not just a generic role filler. There's a fine line between being cocky and blowing their minds 😂
Btw, did you try the game out yet?
The one you pinged me? Not yet. Will eventually.
Yeah that 1 😂 Dementia mode, but in a way that explains how LLMs hallucinate and how cache works
Naaa I am not doing a backflip in CSS just to figure that out bruh. I will play it and let you know later this week lol
All good, lemme know what you think once you do, I'm thinking of maybe making it a series, but depends on if people enjoy them and whether I do the topics justice.
good format for this. the classifier extension is interesting - detecting AI content is a harder problem than it looks from the outside, especially once people start intentionally evading it.
Thanks Mykola! For the classifier, the goal is to be used on dev.to specifically and train on dev.to specifically without relying on LLMs or other AI models in particular. My theory is that if it is more scoped, it will be more accurate and future proof since the detector adapts to the community style of writing. It will be difficult for sure, but I believe we can get there.
the scoped-to-community approach makes sense - general detectors fail on niche writing styles constantly. the part i'd watch is distribution shift as the platform's norms drift. curious what retraining looks like for this.
It would be more manual approach and yes, it's inefficient for now. The plan is having another Chrome Extension where it's goal is to extract data and store it in some database. Then, updating the dataset in the main project.
the extension-as-pipeline is a solid move. manual labeling will probably be the slow part, but that's also where the community-specific signal gets most precise.
Hey Francis, I just checked out "4 Proven Ways To Get Ahead of 99% of Programmers (2026)" from your monthly report, and it actually helped. Thanks for sharing it!
Looking forward to next month's report too 😀
Thanks Hemapriya! Glad it was useful! Stay tuned for more :D
That's a productive month! 🎉
Open-source contributions, AI experimentation, portfolio improvements, and continuous learning through LeetCode are all moving you in the right direction. Looking forward to seeing what you accomplish in July. 🚀
Thanks Anik! Look forward to it :D
How to get into open source contributions?
Hey KV. For me, I would suggest on choosing a repository on GitHub that is a right fit for you based on interests and tech stack preferably. The hardest part is setting up the local environment, but after that, it should be easier to get into it. Hope this helps!
great work
Thanks Akash! :D
😎
😎
You look cute while sleeping!!!!
Thanks! This profile picture was hard to find since there isn't an exact name lol
Some comments may only be visible to logged-in visitors. Sign in to view all comments.