Show HN: I built StickerFacet to turn photos into high quality vinyl stickers | Hacker News Reader

Show HN: I built StickerFacet to turn photos into high quality vinyl stickers

stickerfacet.com

20 pointsarthurcolle14 comments

I decided to focus on pets because my girlfriend and I have this cute little cat called Lola and I immediately needed to produce as many stickers of her as possible.

Eager for feedback! Thank you

Open articleView on HN

I'm not following what the "sticker collection" occupying the prime page real estate at top left is supposed to be. They have names like watercolor, realistic, cartoon, pixel art, neon, and graffiti, but all of them just look like photos. When I click on them there are two buttons with download/upload icons, but neither one is doing anything when I click it.

The "Featured Pet Transformations" below that are good illustrations of the concept, except the "Happy Shiba" example starts from a not-shiba (long floppy ears) and through the magic of GPT-Image-1 becomes an entirely different dog (pointy triangular ears). That makes it look more like an image generator than a restyling of your actual pet.

I wonder if there's anything you can do with the image transformation to make it keep distinctive features of the source material? "Tuxedo Cat" for another example, it makes up a different white and black cat from the one that was provided. The white patch in the original goes up around one eye while the sticker has a pointy triangle up the middle, and its legs were both white but in the sticker they're black with white paws.

EDIT - giving it another look, the "Banana Buddy" example for Cute Style is also turning a bulldog into a terrier

Thank you for the feedback. Those 'transformation examples' are definitely one thing that I was sloppy with. All of the nice outputs are actually derived from pet photos that are legitimately "visibly similar" to the outputs, but I really wanted to push this out so I didn't do my proper diligence.

Here is a video walkthrough to show you some of the meat around image generation / sticker packing: https://www.youtube.com/watch?v=Xcn8dSt7CcQ

I have been working on more advanced features like 1) splitting images into separate images, and 2) re-packing into a single image sheet after validation that the 'biometric color palate' and the 'specific physical attributes that make this thing the thing it is' passes are both sufficiently high in terms of confidence level.

I wanted to initially just have a plain vanilla canvas that you plop images into and then have a "Loom" like experience for regenerating images / creating pseudo-pipelines with just touch/mouse commands and more, but I found that users said it was too opaque and required too much 'obscure prompt engineering experience' to click around to make outputs "pretty" (direct quote).

thank you for taking a look @wlesieutre, I really appreciate it

On the example images, the only one I'd be happy with is the pug (if these were my pets) - the rest actively disincline me from wanting to use the site, as my expectation is now that the site won't do a good job of turning my pet into a sticker.

If I gave those stickers to my kids they would not recognise that they were the source animals.

Maybe just me but the resultant generations look like generic animals. I want _my_ special snowflake little snicker doodle.

I'm a pet lover, just fostered 3 kittens. My brother makes apple messenger stickers of them and puts them all over our group chat. It's great, i love it, they are very very much "hey that's Tilly, look how cute!". vs "it's a cat"

Thank you, this is critical feedback. I totally know what you mean.

Quick question - you tried upload from main page, or did you create a pet in the dashboard and then generate a sticker from that pet page in the dashboard? Not trying to keep posting this video link, but wondering if you were able to get a sticker wall going in the studio: https://www.youtube.com/watch?v=Xcn8dSt7CcQ

ninja edit for myself: Reading back this comment I notice I have too many linguistic abstractions I'm repurposing for this app, but I totally know what you mean @apsurd - and your special snowflake little snicker doodle deserves every variation this construct has to offer.

I'm purely going off the homepage examples, I didn't use the app for myself, given I think it'll just be generic stickers.

A good example is the Tuxedo cat. The original image of the real cat shows the white fur on the face is not symmetrical, there's a cute little dollop of white over the right eye. The generated sticker is completely generic, symmetrical ^ shape of white fur. Real cat doesn't look like that. Not to their owner!

These AI Show HN are just becoming (have become??) boring. The lack of similarity to the original was the first thing that I noticed as well. It just looks like some generic clip art was used rather than genAI type of outputs. It's so bad that I honestly question why the dev would even bother looking for feed back. If you didn't expect people to call out that your output doesn't look like the source, then you're just too scoped in to see the big picture. If the dev did know this but posted anyways, why? It just looks like you have no clue what you're doing. Why why why?

I like the idea, but it looks like orders are public? --> [link-redacted]

has there really only been one?

Oh wow! Full name, phone and address. That’s.. sad for Arthur. I was gonna suggest @dang remove the link but I feel like it almost doesn’t even matter if it’s that easy to find

Needs tuning because the examples don't look like the original pets.

The examples definitely need to be changed, but I am considering just throwing away this implementation and restarting as I don't want to be tied to this current typescript implementation and I think I can move a bit faster with Rails.

The underlying pet generation is actual not too bad, but you're fundamentally right -> need to extract features from the core pet content and then use those to generate the sticker assets. I already do this - in fact I have a very advanced pipeline to take 1) OG image -> 2) create 30-50 variants -> 3) compare against stock to get good 8-15 variants -> 4) take those and generate prompts/images that are useful in creating the "final good stock"

From those, if there are multiple images on any single image, I split them. Then, as the user uses the app, the images get packed into one or multiple sheets that are then able to be shipped.

Here are some of the original animals for the sticker examples:

https://x.com/arthurcolle/status/1926469831970165097

I'll fix the app examples. Thanks for the shout.

Like the idea but is any of this actually working? Created an account, uploaded a photo of my dog. This bring me to a page which "Sticker Collection" which is showing me a bunch of cartoon cats, which based on their title are supposed to be different styles but all look the same to me. It's not even showing the original photo I uploaded. It has an "Add to cart" button when I select one of the "styles".

No idea what I am supposed to do with this.

The pricing link in the menu does not work. I could not find what sizes and material for the physical stickers you used. "high quality" hmm

Reply on news.ycombinator.com