I sent robot forgeries to a handwriting expert

I cant believe you managed to create machine learning code for doctors handwriting on the first try

What I learned from this channel over the years is that in order to do less work you have to do more work than you originally had to.

I love the "wife annoyed to be forced to help her husbands weird projects" character she pulls lmao

I love the chemistry between him and his wife. They have the same sense of humor and banter so well. Ugh.

This video is the embodiment of "we do things not because they are easy, but because we thought they were"

Ok, one major tip: natural hand writing is in fact a 3D action not just 2D, meaning that the writer exerts higher and lower pressure vertical to the paper surface as they write, which results in the pen line becoming thinner and thicker at different sections of a letter! Next try to build the Y-axis movement into that robot!!

I would like to thank you on behalf of all criminals for giving us starting point of forgery and also explaining us how we might get busted so we could fix this before we go live.

I feel like, in this case, a forensic handwriting expert being able to make a profile for your handwriting bot is a feature, not a bug. After all, you're not trying to make 20,000 unique sets of handwriting, you are trying to make one set of handwriting that is consistent across 20,000 use instances. And if he sees enough shared characteristics between the different pieces of writing to work out a profile, that means they are identifiable as having been written by the same person (or, I guess, robot). Which means that you succeeded in creating a unique and consistent handwriting style

I'm surprised pen pressure on the paper wasn't more of a problem. Seems like the robots perfect line darkness would stand out more.

just a tip when using neural networks. In the video, I noticed after every bug you fixed, the editing at least made it look like you spent ~50 hours training the RNN again. Usually, you can use smaller datasets to train the networks and see if the output is slightly acceptable before spending the 2 days training the network with the full dataset.

I love his wife’s facial expressions. It’s just the look of someone who loves a benign lunatic genius.

the most educational part of this is when he says, "on the first try, too! that never happens!" I always got discouraged as a kid when I didn't get things on the first try and i gave up. i didn't have any confidence to try again because i always tried my hardest the first time. If my best try wasn't good enough, no further tries seemed like they'd fare any better, so i, being a very reasonable and smart kiddo, concluded i just wasn't very good at that thing.

I love the field of Computer Science. Spending 4 months to create something to do a 3 hour task for me just gives such a huge feeling of accomplishment.

Trained as a mechanical engineer 40 years ago - despite afterwards working in another field your videos resonate with the engineer's heart that still beats within. Thank you!

I can’t imagine how much time you put to craft those awesome videos! Amazing!

Absolutely love how succinctly you managed to sum up the experience of learning machine learning: write code; wait hours/days; find out you made a really dumb mistake; repeat steps until you eventually either ragequit or swallow your pride and decide to see if someone way smarter than you already figured it out (SPOILER ALERT: they did). Actually I suppose this goes for a lot of things!

The shot @9:37 has me holding my sides. A $35k robot arm, TWO computers, a big power cabinet, an air compressor, a shop-vac... "But that would be over-engineered!" 😆

You know you're an engineer when you spend hundreds of hours designing and building a custom solution to do a simple menial task

This is probably the best visual description of gradient descent I've seen! Awesome video!!

A very very few people in the world got what it takes to produce videos like this. Technical and theoretical knowledge, a good sense of humor, and video editing skills. this man deserves a medal!