Show HN: Crowdsourced Object Recognition

http://www.crowdsend.com/

by argumentum • 14 years ago

77 32 14 years ago

timmyd13• 14 years ago

Hey Guys,

Thanks for checking out our product. Been working hard on it to resolve the problem of object recognition and trying to produce something thats different to what's out there right now.

If you know anyone that want's to help or get involved - love to speak to them or if you want an invite just let me know and I'll send it right now.

Love your feedback on http://about.crowdsend.com too if you can! Open to anything.

argumentum• 14 years ago

Btw, Tim's from Australia and has been staying at my house in Los Altos for a month or so. Gotten to know him well, great guy and fantastic front-end hacker.

We've been working together putting some finishing touches on CrowdSend. I think it's going to be a winner.

jermaink• 14 years ago

good luck, mate! Loving to be behind the scenes :)

caublestone• 14 years ago

Great idea. Make a mobile app with an incentive so people can ID images to waste time. Focus on fine tuning an algorithm by checking against what people say and we could be on our way to making a really good image processor.

timmyd13• 14 years ago

yeah thats in the works at the moment mate. great idea with the algorithm

we're totally bootstrapped so just trying to get the web app sorted at the moment and then we'll get that out!

ewang12• 14 years ago

you could do a game. i've seen people getting really addicted to those iOS logo ID games.

for the incentives, just use something like kiip.

dm8• 14 years ago

Congrats on launch! Not to spoil your party, there is another startup that is working in the similar space - www.luminate.com

What will be the primary use case of your product?

timmyd13• 14 years ago

What we solve - which we think is the biggest problem - is being able to ask what is inside an image.

You as a user, can grab any image, from any website - and ask what is inside it and others can help you. That's what we believe is the biggest problem.

The problem isn't arriving at an image and knowing - its arriving and not knowing. Luminate (and others) require the publishers to tag - however what if you arrive at some website (like the entire of instagram as one tiny example) and want to know what's in an image - where do you go ? Google - nope. That's just similar images they don't tell you any more. Luminate won't help you either.

So that's the difference - luminate and everyone else doesn't actually help you solve a problem as you surf the web.

Love to give you a preview if you're interested - we've done all this bootstrapped. We are only in beta so there will be bugs!

greendestiny3• 14 years ago

Not to mention rolling over something in luminate pops up a lightbox that obscures the image and has to be closed.

seagreen2• 14 years ago

This is a really cool idea and I'm glad someone's doing it. Have you considered talking to Amazon about it? It's definitely in their best interest for products to be easily identifiable, and they're already doing work on it in a different way with Price Check[1].

Unfortunately there are other companies with a strong incentive to make identifying specific products as difficult as possible[2]. And clothing companies have never been good about sticking with a specific product anyway. I got a Ralph Lauren suit last year that I liked enough to consider ordering another one, but they had already stopped making that style. Maybe dresses or something I could understand, but men's suits?

If you can't buy a product even if you can identify it, then what's the point? Do you have any thoughts on how to get around this problem? One idea that comes to mind is that you could use your data to build a ranking of how long clothing companies (or any companies) stick with their products. I'd certainly find it useful.

Best of luck!

[1] https://play.google.com/store/apps/details?id=com.amazon.pri...

[2] http://online.wsj.com/article/SB1000142405297020462420457717...

timmyd13• 14 years ago

thanks for the comment mate!

yeah - we currently actually work with a heap of retailers and we do match to where you buy them - we've around 90 or so at the moment (love to get your thoughts on http://about.crowdsend.com).

publishers can do a bunch of stuff too - but we think the bigger problem rests with the community than with publishers.

basically once a product's been identified then we match it to a bunch of retailers if we can find it!

greendestiny3• 14 years ago

Love the idea, sent through a beta invite request. Not sure about the fade in on the images on about.crowdsend I find it a bit annoying combined with scrolling - its like having two transitions on a powerpoint slide.

seagreen2• 14 years ago

That seems like a great way to go. I always felt somewhat embarrassed when googling bar codes trying to find a shirt or whatever. As a programmer I knew there's a better way for this to be done, which made the whole experience even less pleasant than it was already (and frankly it was already pretty boring). Glad to see you're stepping up to the plate.

ewang12• 14 years ago

even if you can't buy it, the tagged image data is going to be pretty valuable... couple it with an image matching engine, and they can charge for API access.

driverdan• 14 years ago

As a (mostly) non-consumer myself using this doesn't really interest me but I can see how the Pinterest crowd would love it as well as the businesses who make the products.

Do you automate some of the product identification or is it only by users?

timmyd13• 14 years ago

At the moment - we are just focusing on getting data together from users which has been a fair bit of work. We do some automation in the background - but we're bootstrapped so its all what time we have!

We think being able to ask what something is and getting a response from a community is something thats a bit different in the image space.

frozenport• 14 years ago

I dragged a box over the ipod demo image and tried to identify as an ipod. It told me it couldn't find anything called "ipod"?

timmyd13• 14 years ago

yeah sorry - the server is getting smashed at the moment - have to free up some mem!

but let me know what browser so i can double check ?

mef• 14 years ago

Interesting product. What's the incentive for someone to do the research to figure out what's in an image?

timmyd13• 14 years ago

yeah at the moment we are doing a point based system where user gets discounts from retailers. unsure what will work best but its something that we going to play around with until we find what works best

kwekly• 14 years ago

I deal a bit with arty types, and they tend to pride themselves on being able to identify products, along with designers and quoting how many hundred $'s something "costs"

I think it's a bit like nerds and bragging rights in being able to produce good explanations for difficult technical questions. Along that line of thinking, perhaps a stackoverflow karma / community recognition style system might be enough.

hayksaakian• 14 years ago

I really Google doing something like this a long while ago. They turned it into a game of some mind. (My memory fails me)

timmyd13• 14 years ago

Yeah but we are a bit different in our approach and the way we are trying to resolve it :)

neya• 14 years ago

Really cool concept mate, congrats! Keep up the good work. One small feedback - The photo of the guy on the frontpage looks more like Justin Beiber (I thought it was him, for a second) :D Other than that, everything else is perfect!

Cheers, Good luck!

timmyd13• 14 years ago

thanks mate :) been a lot of work to get here and its only the start!

keshansodi• 14 years ago

Good luck men, great idea.

timmyd13• 14 years ago

thanks a heap. if you're after an invite just let us know

tled• 14 years ago

Awesome man, I would like to integrate it on my Website and my iOS/Android app. Hope you guys open the API soon. <3

Show HN: Crowdsourced Object Recognition

Related Stories