so it's ok to say "SSD read/write speed", but now that we have something closer to the original meaning of the word, someone always has to point out that "LLMs don't have a soul" (or whatever you think is required for it to count as akchyually reading)
If I can just stand up for the nitpicker - arguably in the uncanny valley it’s more natural to point out it’s not reading (by their definition) than outside it (ssd’s).
makes sense in a philosophical debate or when you're talking to your confused grandparents, but does anyone on hn not know how LLMs work, at least on the level of "tokens, matrices, data, sgd"?
otherwise, that reminder must imply that people do know how it works, and yet they still ascribe to these models some property like qualia, i.e. something other than "being able to turn english into code and compute into shareholder value";
but then if you disagree, why even mention it in the first place? do atheists randomly proclaim "btw god isn't real!" in unrelated conversations with strangers of unknown religious beliefs?
Hey, I'm cofounder.
Overall relevant summary, but here is extra context:
ai;dr - neither of us are english natives. We use AI, but the content and research is ours and AI could not do it well enough. It's n'th iteration. In fact, we started with photo-based approach and over time simplified it to single questionnaire which yields better results. simple != easy != obvious
why: no fancy equipment, no weird camera scanning (yielding bad results anyways). Still plenty of opportunities to make it even simpler. Ideally we'd measure full body as well or better than a tailor with a tape measure in just few minutes. Possibly without tape measure. We're still long way off both in tech and UX, but good enough for market validation
errors:
all of that is roughly within the errors that you'd get from tape measurements. If it's could enough for bespoke, it's good enough to tell you that your bicep won't feel comfortable inthe sleeve
Twenty or so years ago, Levis had a program called Personal Pair (they paired with a firm/product called Intellifit) where you would step into a millimeter wave 3D scanner and get your precise measurements.
Which you could use to get a custom pair of jeans made just for you. This failed partly because the target market turned out to be middle-aged adults (that buy fewer pairs), and not 18-25 year olds like they were hoping. Partly over privacy concerns. Partly the ability of the factory to make jeans to that tight of a tolerance. And partly because it was promoted like it was a novelty ("Consumers love it!"). Mentioned here previously:
But with this software - the tolerances are looser, so the clothing becomes more manufacturable. And the measurements can be anonymous - you don't feel like you're stepping into a TSA scanner for everyone to see.
I hope they are able to make relationships with multiple clothing brands so shopping from home will become less hit-or-miss. The benefit to the brands is going to be fewer returns for size issues.
There should still be privacy concerns, especially with their demo which sends a POST on "Generate". The author suggests the model is 85kB of weights, which could run perfectly well in browser.
> But with this software - the tolerances are looser, so the clothing becomes more manufacturable.
Does it? How do looser measurements help? I assume manufacturer would always take the upper bound of dimensions. Suppose model also predicted your dimensions are higher then they really are, so these two in combination give you an oversized piece of clothing.
Not just oversized - undersized also happens. Most cloth is still cut by hand using large electric saws and it's just not that accurate. (caution: loud music)
Notice that the panels are marked out with chalk and if the operator doesn't stay square to the table, or isn't diligent in marking up the panels, they won't be consistent with the brand's standard sizing.
I mean - ideally a set of panels of a piece of clothing would be cut by computerized laser so it's accurate to what the buyer needs. But that costs too much and takes too long.
Thanks, this matches what we're seeing. Trying to establish relationship with some brands, this is the problem we hit. The clothing batches can be very different from each other - cut-by-hand problem. So they don't have sewing patterns to drape and even if they have the final clothes can be different. Plus they don't want to share the patterns/technical details.
So on business side despite the clear benefits, for now we have hard time finding interested brands. Probably part of it is that we're very technical, technology-focused guys. But we're evaluating both paths: whether the mass-made item will fit, and tailoring for a specific person. Will see how it works out.
On the scanner side. The software approach beyond the less friction also have a benefit of predicting the future shape: "pregnant me in 2 months" or "me with 3kg less". Or simpler: my measurements changed since last month and I don't need to rescan. That's harder with hardware.
So perhaps you need to present them with a more complete solution. Develop or source a faster laser cutter. And find a small factory in an area with skilled labor (central North Carolina still has people who can cut & sew). And develop a process in the factory to keep all the panels for one item of clothing together from cutting to sewing to shipping (maybe a bucket?).
It'll be a bespoke item at premium prices, no question. How big is the market for this? I don't know, but my feeling is that it's older women.
(Cofounder here)
I talked to really big fast fashion brands on that subject.
Oftentimes they dont even have the garment cuts at hand. They sort of outsource big part of the designing process to the supplier (Bangladesh, China contractor).
Fashion Technicians are handed samples and they iterate on them, but they can't reproduce the actual fabric dimensions that easily.
The fast fashiuon brands are very hard to get them to experiment with such solutions for that matter - it would interfere with very fast and scrappy process thats in place
I don't understand why the height and weight errors aren't 0 when they are known inputs? If I say how tall I am, why is the model estimating something else?
That's a common phenomenon in model fitting, depending on the type of model. In both old school regression and neural networks, the fitted model does not distinguish between specific training examples and other inputs. So specific input-output pairs from the training data don't get special privilege. In fact it's often a good thing that models don't just memorize inputt-output pairs from training, because that allows them to smooth over uncaptured sources of variation such as people all being slightly different as well as measurement error.
In this case they had to customize the model fitting to try to get the error closer to zero specifically on those attributes.
Yes, but why are they estimating the features when they are already available? They can estimate the other measurements from height etc, and just use the known inputs as is. I don't get the point of passing them through a model at all.
The previous response was exactly right. The estimated features are impacting height, so the height can't be set then do the rest. It also cannot be tuned afterwards because it would change the mass. So vicious circle.
It takes more like 10 seconds. For a large range of height and weight inputs crossed with all option combinations, you could precompute ~10M measurements and return results basically instantly.
Yeah, the demo wasn't prepared for such peak. Normally it's <2s after warm-up. Like the precomputation idea, but for now it changes too dynamically to precompute each time.
Wrong. The article is our own research. Writing is claude assited, cuz we are not english natives. As for the code it's also Claude, but once again, LLMs are surprisingly bad at 3D ATM
Interesting idea. Using a questionnaire as input for an MLP makes sense but the real challenge is designing questions that capture useful signal instead of noise. If that part is done well, the approach has a lot of potential.
For the humour impaired, pocketsize in women's clothing is contentious because for reasons determined by the fashion police pockets are unsightly and either fictional, or tiny.
One time, I found women's jeans with tiny pockets but the bottom seam could be pulled undone to reveal a man size pocket of cloth below. That was classy.
My guess, the article itself is clearly AI authored and there are a fair number of us who don't particularly like the writing style. Further, it implies something about the original human's own valuation of this work - if they decided to let the machine handle it, why should I spend my own time reading what they didn't bother to write?
I don't like AI slop either, but we're both non-natives, so we did a run of AI after initial draft for the reader's comfort. We'll try to improve on the style
I'm guessing the writing is AI-assisted (there's no fluidity and it has some weirdly placed phrases) but I see they're in Poland and likely not English-language first?
MLP trained on 8 questions achieves ~0.3cm height error, ~0.3kg weight error, and ~3-4cm for bust/waist/hips measurements.
https://www.mdpi.com/1424-8220/22/5/1885 + some hacking => "we want to productize this"