This post has been de-listed
It is no longer included in search results and normal feeds (front page, hot posts, subreddit posts, etc). It remains visible only via the author's post history.
After playing around with my AI coin grading project for half a year, I've started to sour on the idea. I think Technical Grading will eventually be accurate enough to be useful, but I don't think the market will accept it. PCGS, NGC, and CAC are just too dominant. People want and trust the judgment of these companies. I just can't see AI grading taking off in a big manner.
But...after testing AI over a thousand times, I noticed that it's extremely accurate in identifying coins 99% of the time. Specifically identifying the series, year, and mintmark. [For US and World Coins, not medals or tokens. Nor varieties/errors]
I've been working towards a robotics system to identify and sort coins using ROS2, a Raspberry Pi, a mini camera for the AI to take and analyze images, and Legos bricks. Consider it Numi v2.
The goal is to build a coin-sorting robot that uses AI to intelligently sort coins. Imagine a reverse Coinstar that spits back out valuable coins.
My friends working at coin shops lament about all the bulk coins coming in that they don't have time to sort. Such as wheat cents and silver world coins. Which is a bummer because they're leaving a bunch of money on the table.
I'm starting with wheat cents, but the goal is to work towards sorting other US coins, silver world coins, and eventually identifying varieties & errors.
Attached are images of the initial designs I built using Lego Studio Designer. I'll be building the physical prototype over the next few weeks.
I'm looking for feedback and suggestions.
What would you want out of a coin roll hunting machine? How could it machine help you or your coin shop?
![AI-Powered Coin Roll Hunting [Project] image](https://enm-prod-tns.sfo3.cdn.digitaloceanspaces.com/95c6acc7d6e30f64059491a94fd0d7aa-md.png)
Image seen in 1 other post by this user.
![AI-Powered Coin Roll Hunting [Project] image](https://enm-prod-tns.sfo3.cdn.digitaloceanspaces.com/05b9cbfd3f777199bbe4c560c534620a-md.png)
Image seen in 1 other post by this user.
Reposted Media Detected
One or more of the images in this post exist in 1 other post
Subreddit
Post Details
- Posted
- 5 months ago
- Reddit URL
- View post on reddit.com
- External URL
- reddit.com/gallery/1dd4w...
"What will you use for training data?"
I'll be using the GPT-4o model so no customization. I used GPT-4 to power my AI grading app, Numi.
https://www.reddit.com/r/coins/comments/1c9qagd/revisiting_numi_testing_the_latest_gpt4_update/
While it was a miss on AI grading, the model performed spectacularly in correctly identifying US and world coins.
"Would it split things out as "Junk", "Great", "Fake", and "Heck if I know"?"
The first focus is on sorting wheat cents by rare dates and by decades. Anything it can't determine will be put into a "sort by hand" bin. The GPT-4o model isn't good enough to detect counterfeits. I'd like to imagine letting it run all night and waking up to find a 1909-s VBD :)
Long term the hope is to train my own AI model to detect errors and varieties.
"I'm not sure how much you could add to it, but something that also does a coin weight to help identify fakes might be an addition section."
I've heard this suggestion before and it's interesting to imagine a sorter that can also weigh a coin. Wayyy long term I'd like to explore detecting counterfeits, but right now the model isn't there yet.