Hacker News: jbaudanza

New comment by jbaudanza in "Show HN: Happy Hangul Day! An RNN for Generating Korean Handwriting Strokes"

jbaudanza — Sun, 11 Oct 2026 16:00:01 +0000

That's an interesting idea. Velocity is something that can be derived from the x, y inputs that the model already has. So, if the velocity contained useful information, I would expect the model to be able to learn this during the training phase. The pressure, on the other hand, was new information that the model didn't have before.

New comment by jbaudanza in "Show HN: Happy Hangul Day! An RNN for Generating Korean Handwriting Strokes"

jbaudanza — Sun, 11 Oct 2026 14:23:06 +0000

The text embedding structure is very specific to Hangul. Someone that knows more about Chinese than me could probably design a suitable text embedding structure that would work well. I’m happy to collaborate if you want to fork my project and try.

New comment by jbaudanza in "Show HN: Happy Hangul Day! An RNN for Generating Korean Handwriting Strokes"

jbaudanza — Sun, 11 Oct 2026 14:17:39 +0000

Thank you for the feedback. I’ll incorporate this into the new training run.

I’m sorry about your father. Reproducing his handwriting would be a beautiful project. What kind of samples do you have from him?

New comment by jbaudanza in "Show HN: Happy Hangul Day! An RNN for Generating Korean Handwriting Strokes"

jbaudanza — Sun, 11 Oct 2026 10:58:52 +0000

Oh thank you, good catch. "쀍" is definitely not a common character in my training set. I'll add it to the next run!

New comment by jbaudanza in "Show HN: Happy Hangul Day! An RNN for Generating Korean Handwriting Strokes"

jbaudanza — Sun, 11 Oct 2026 06:48:44 +0000

Ah yes, I'm well aware! My company (hiworker.com) makes AI products for ship building and construction companies that have a large foreign workforce. We have some customers in Ulsan. It's a beautiful place, I love visiting.

New comment by jbaudanza in "Show HN: Happy Hangul Day! An RNN for Generating Korean Handwriting Strokes"

jbaudanza — Sun, 11 Oct 2026 06:30:17 +0000

Thank you, you're right! It definitely seems to have a problem when the end of the string ends with 다. It seems to work if you end the sentence with "다." instead of "다". I'll look into this more. There should be plenty of "다" data in the training set, maybe there's something deeper going on.

PS: I saw your feedback report. 화이팅! :)

New comment by jbaudanza in "Show HN: Happy Hangul Day! An RNN for Generating Korean Handwriting Strokes"

jbaudanza — Sun, 11 Oct 2026 05:01:50 +0000

I think this would work well for Chinese, but you’d probably need to modify the text embedding system. Let me know if you’d like to try, happy to collaborate.

New comment by jbaudanza in "Show HN: Happy Hangul Day! An RNN for Generating Korean Handwriting Strokes"

jbaudanza — Sat, 10 Oct 2026 04:28:18 +0000

> Nice! Did you find any weird/funny case while training the model? [I remember autotranlation/readers inventing stuff a few years ago.]

Mostly, it was me trying things the naive way, running into problems, and then solving them, usually by copying what Alex Graves did in 2013 (but not always!)

For example, to make things simple I first built the model to simply predict the next x, y direction of the pen. This worked for simple pen strokes, but I noticed the model had a difficult time turning corners.

To fix this, I changed the model to instead predict an Mixture Density Network (Same as Graves). This is explained better in his paper, but essentially, instead of predicting one x, y direction, you predict 10, and then randomly sample one of those predictions. Also instead of predicting scalar x, y values, you predict a parameters for a gaussian distribution, and then sample from that.

It always amazes me how much randomness is involved in intelligence.

> If you get someone to make a few samples and photos, it would be very nice.

I'm in Seoul, if you know anyone who would be interested in collaborating, please send me a note! jon@jonb.org

Show HN: Happy Hangul Day! An RNN for Generating Korean Handwriting Strokes

jbaudanza — Fri, 09 Oct 2026 03:58:16 +0000

Article URL: https://hangul.ink/blog/hangul-day

Comments URL: https://news.ycombinator.com/item?id=50015833

Points: 88

# Comments: 16

Show HN: Node.js native module for integrating Mediasoup with Audio AI models

jbaudanza — Mon, 16 Feb 2026 00:47:24 +0000

Article URL: https://github.com/Hilokal/audio-rtp-tools

Comments URL: https://news.ycombinator.com/item?id=47029498

Points: 1

# Comments: 0

New comment by jbaudanza in "Fuckarounditis (2011)"

jbaudanza — Wed, 26 Mar 2025 00:07:02 +0000

I had the same problem. I think a lot of people reading this will have weak hamstrings. I think it's a natural consequence of desk work.

Adding in some Romanian Deadlifts really forced me to focus on my hamstrings when doing lifts. It helped me breakthrough some plateaus with the conventional deadlift.

New comment by jbaudanza in "OpenAI Audio Models"

jbaudanza — Fri, 21 Mar 2025 00:56:41 +0000

3) Whisper really needs to be paired with Silero VAD, otherwise the hallucination problem makes it almost unusable.

New comment by jbaudanza in "Auntie PDF – an open source app built using Mistral OCR"

jbaudanza — Sat, 08 Mar 2025 05:21:38 +0000

I have a question about Mistral OCR. If I give the model a PDF that is 90% text, is it actually performing OCR on an image representation of the text? Or is it smart enough to extract the text directly and only use OCR on images?

New comment by jbaudanza in "Show HN: Ricotta – Language Learning to Replace Anki"

jbaudanza — Mon, 10 Feb 2025 02:26:55 +0000

I really like this concept, and am working on a similar product for my own self-study.

I would like to see more language support. For me, I'd like to study Korean

The UX around card navigation needs some work. It took me a long time to figure out that after I chose a smiley face, that I was moving to a new card. This was especially true if the next card is similar to a previous card.

New comment by jbaudanza in "OpenAI DevDay 2024 live blog"

jbaudanza — Wed, 02 Oct 2024 01:24:25 +0000

Interesting choice of a 24kHz sample rate for PCM audio. I wonder if the model was trained on 24kHz audio, rather than the usual 8/16kHz for ML models.

BlockJoy raises $12M cut operating costs for businesses running blockchain nodes

jbaudanza — Fri, 03 Feb 2023 19:16:05 +0000

Article URL: https://techcrunch.com/2023/02/02/blockjoy-raises-12m-to-help-cut-operating-costs-for-businesses-running-blockchain-nodes/

Comments URL: https://news.ycombinator.com/item?id=34646421

Points: 4

# Comments: 0

New comment by jbaudanza in "Ask HN: Advice that changed your life?"

jbaudanza — Sat, 29 Oct 2022 09:38:28 +0000

Righty tighty, lefty loosey

New comment by jbaudanza in "Enhanced noise suppression in Jitsi Meet"

jbaudanza — Sun, 02 Oct 2022 00:43:24 +0000

This looks interesting.

Most devices have noise suppression built in. Do you use RNNoise in tandem with the hardware noise suppression, or do you disable it?

Also, are there any plans to implement this with react-native on mobile? And if so, how would you implement this, since audio worklets and WebAudio aren't available.

New comment by jbaudanza in "Ask HN: What lifestyle choices have led to improved cognitive function/memory?"

jbaudanza — Wed, 29 Jun 2022 10:19:32 +0000

Yes! I’ve had the same experience with spaced-repetition. I’ve used Memrise.com to learn a large set of Korean vocab words.

I’ve found it’s improved my general recall ability, not just for things I’ve scheduled into a spaced-repetition program.

It’s like the program trains my brain on how to recall things, and then I can use that skill on anything.

I’ve also found that if I slack off my training schedule, my general recall ability gets worse too.

Also, alcohol for sure impacts my memory. I find even one drink will make it harder to recall things

New comment by jbaudanza in "Flutter is better than React Native in all the ways that don’t matter"

jbaudanza — Wed, 16 Mar 2022 19:02:06 +0000

I know this is theoretically true, but I've never found this to be the case in practice.

Coinbase, in fact, has made it a rule that all things must be put in useMemo. https://attardi.org/why-we-memo-all-the-things/

Facebook is working on an experimental compiler called React Forget that will transparently memoize everything for you.

https://www.reddit.com/r/reactjs/comments/rcn5ks/react_forge...