folli

Can you recommend any kits? Out sounds like an extremely fun hobby.

Reminds me of this classic: https://youtu.be/tVo_wkxH9dU

Please don't. The project is around since at least 2022.

Spending years on a cool project, and than some dude posts such a comment...

You can use an IDE like IntelliJ and you get syntax highlighting, code completion etc.

I think it can be more efficient for two-step RAG so you can reuse the natural language query directly, but for agentic RAG it might indeed be overkill.

I was just working on a RAG implementation for >500k news articles, completely local, using postgres as a vector database: https://github.com/r-follador/TeletextSignals

I'm positively surprised on how well it works, especially if you also connect it to an LLM.

Less ichy than cotton

Also, depending on the ingredient, it makes more sense to use cups as a measurement of volume, not mass, when converting to metric. E.g. liquids, yoghurt etc.

Another thing: although not strictly metric, but European recipes also use tablespoon and teaspoon as measurements for smaller volumes, so no need to convert this.

Just my two cents, other than that very nice work!!

Depends: for hobby purposes, or what my daytime job forces on me?

Extremely tight results and the majority of cantons voted against it.

I still believe it's the way forward.

As a hobby project, I was looking into using LiDAR data to view archeological points of interest in Switzerland: https://github.com/r-follador/delta-relief

It would be interesting to overlay TESSERA data there, although the resolution is of course very different.

How do you use it with Jetbrains? Junie? Or just as a separate CLI session?

This argument is not limited to gene therapies, but would apply to pretty much every pharmaceutical product.

One issue is however that the actual costs are not so much in early R&D (what the publicly funded universities and hospitals are doing), but in the later stage (clinical trials) which needs deep pockets and appetite for risk, which only big pharma has, because they see a potential big payout.

Tangentially related: I'm trying to make LiDAR data in Switzerland more accessible, see https://github.com/r-follador/delta-relief

There's some interesting examples in the Readme.

Why ice?

SwissTopo has a separate dataset of buildings and structures in Switzerland, so they basically just subtract it from the LiDAR data.

Here's another article about the use of such data in South America: https://www.nationalgeographic.com/history/article/maya-lase...

Of course, nothing so exciting to be discovered in Switzerland anymore ;)

I think the examples should make it more clear. Thanks to the high resolution of the data, you can see subtle changes in the slope (aka relief aka microtopography) that could hint to underlying remains of human settlements (usually some suspicious geometric patterns that you would not expect in a natural terrain).

See also here for an in-depth discussion on the potential use of such data: https://www.mdpi.com/2072-4292/15/6/1569

How do you suggest to change the description to make it less confusing?

I can't get it to work, but it sounds interesting. Does anyone have a screen recording to share?

Do you have any links further elaborating this?

Any recommendations regarding freight forwarders?

Hijacking this thread: what's currently the cheapest way to get structured data out of a PDF?

I assume there's some reasonable tool out there to convert PDFs to Markup and than feed it to some LLM API with okay costs (Gemini? DeepSeek?). Any suggestions?

Fighter jets emit ADS-B? I know it's only training, but I'm curious?

So in "Paskalampi" and "Vesijärvi", which part of it means "lake"? Naively (I have obviously no clue about finnish), I would have expected some syllables (or token in modern LLM terminology) to occur in both names (i.e. representing the "lake" part).

Nice!

I'm working on something similar but aimed at visualizing GPS tracks, e.g. for hiking and biking: https://cubetrek.com/view/6638

Let's share some notes, if you're interested! Code is open source: https://github.com/r-follador/CubeTrek/

This is fascinating, didn't know. Why does higher demand lower the frequency?

Looks great!

Anything you can share? Papers? Examples?

Has anyone worked on AI/ML approaches for detecting ancient structures in LiDAR data? Given that training data for identifying remnants of man-made structures (e.g., Roman or medieval ruins) is quite sparse, how would you approach this problem?

Some initial thoughts:

- Data Augmentation: Using synthetic or simulated LiDAR data from known structures to improve training. - Few-Shot Learning / Transfer Learning: Training models on better-documented archaeological sites and applying them to new areas.

Would love to hear thoughts from people with experience in remote sensing, computer vision, or archaeology!