Excellent work guys! If you can make your service cheaper and keep your promises, you have a winner.
Show HN: We built an end-to-end encrypted alternative to Google Photos
411–420 of 421 posts
Re: Show HN: We built an end-to-end encrypted alternative to Google Photos
#412Earlier quoted context omitted.
Datasets are actually pretty hard to create. You can see several papers specifically studying ImageNet[0] including some on fairness and how labels matter. There's also Google's famous private JFT-300M dataset[1]. JFT was specifically made with heavy tails in the distribution to better help study these areas, which is specifically the problem we're interested with here and one that is not solved in ML. Even with more…
godelski, I really appreciate such a thoughtful response to my curiosity. Looking at this while better understanding the problem, I wonder what features I really want for my own photo library. Thinking of iOS photos. Matching people together seems hard. But grouping photos by GPS location or date is trivial. So we have to get clear on what features are important for home photo libraries. I can now see how the idea of…
Hey, that's what HN is about. You got experts in very specific niches and we should be able to talk to each other in detail, right? That's the advantage of this place as opposed to somewhere like Reddit. Though expanding size we face similar issues.
These are good points about GPS and other metadata. I didn't really think about that when thinking about this problem, but every album I create is pretty much a combination of GPS and temporally based (though I create this with friends). But I think you're right in suggesting that there are likely _simple_ ways to group some things that aren't currently being done.
> I can now see how the idea of "use public libraries = solution" falls short. It neither presents a viable solution or demonstrates rigorous understanding.
ML is hard. But everyone sells it as easy. But then again, if it was easy why would Google and Facebook pay such a high rate for researchers? There's a lot of people in this space and so it is noisy. But I think if you have a pretty strong math background you start to be able to pick out the signal from the noise better and see that there is a lot more to the research than getting SOTA results on benchmark datasets.
Re: Show HN: We built an end-to-end encrypted alternative to Google Photos
#413Earlier quoted context omitted.
I'm in the same boat, have been watching, love that they have a businessmodel and am waiting for the time when they are covering my needs (face recognition, object / scene detection...). I'd even pay a 2$/month "lurker" subscription which has like 100mb of storage so I can check the features from time to time and support the team.
As someone who's never used cloud-based photo browsers... I always assumed the facial recognition aspect was primarily for social media apps that try to tag known faces from a user's friends group, to put it in those people's news feeds or something. It's one reason I avoid being photographed and ask people not to tag my name to my face if they do post a photo I'm in. I'm wondering, what's the utility of facial recog…
It's so incredibly useful to be able to bring up pictures but you don't remember the exact time or date that you took it.
Google photos has come in so clutch when you're searching through 50k photos.
Re: Show HN: We built an end-to-end encrypted alternative to Google Photos
#414Earlier quoted context omitted.
Hey, so the project had initially started off as a self-hostable software (with an option to buy a pre-configured device). We realized soon that it's hard to monetize such a product in the consumer space to the point where it can become self-sustaining. We don't have a problem with offering a self-hosted variant. But given our limited engineering bandwidth we had to take a call on who our target market should be, and…
> We realized soon that it's hard to monetize such [self-hosted product] Spot on. We iterated on a similar product in this space: "privacy preserving", "self-hosted", "open source" etc. But focused on local AI indexing & search of personal videos and photos [0], rather than backups. We ultimately shelved VideoNinja because we weren't able to find a sustainable business angle: * Non-technical people simply don't care…
While it unfortunately didn't work in the consumer market, there's a space for video recognition in the business space:
- Scene finding for directors/news channels. AP and other sources have a lot of material but you pretty much literally have to watch the entire video in order to find a good scene.
- Scene finding for the XXX crowd. Very underserved market.
- Scene finding for police/lawyers. While it may seem like the opposite of 'privacy preserving', defense attorneys are literally just swamped with video evidence in an attempt to make them give up. Similarly if you're suing a big company for something as simple as an on the job injury or harassment, and need to prove there's a pattern of harm... they'll give you everything and let you do the work of finding out that there was a pattern of bad behavior.
It's the kind of thing that'd be useful as an open source solution... or failing that having a company which is 100% neutral in operation is also good.
I'm currently using Microsoft for something like this because they're absolution massive and apart from their OpenAI division, they only care that what you process is legal.
Re: Show HN: We built an end-to-end encrypted alternative to Google Photos
#415Earlier quoted context omitted.
Ah, what you really meant was "what's evil about selling my data?" which is a much larger question. And it sounds like you already have your answer.
They actually don't "sell".
They take your data and turn it into something that has value to them. With actual selling, that something is money. In this case it could be something else, but saying it that way will not help the general discourse of this problem at all. Much like being pedantic over terminology.
Re: Show HN: We built an end-to-end encrypted alternative to Google Photos
#416Earlier quoted context omitted.
That is not categorically true. On the business side, there's plenty of companies that have offered and succeeded with self-hosted software. On the client side, there's many individuals like myself willing to dedicate time, money, and effort to self-host services. I spent quite a bit of time setting up my NAS with self-hosted services, not only because the number of photos and media I store would be prohibitively exp…
we have so many consumer facing apps. you'd want to maintain all those and actually have a life to use those? good luck!
Not everybody has to use "many" apps. You can only self-host those you care about.
Re: Show HN: We built an end-to-end encrypted alternative to Google Photos
#417Earlier quoted context omitted.
Just put a price on it, ffs! Make it extensible with plugins. To gain 100% trust make it open source. I am happy to pay good money of a local, non-leaking AI based tagging software for video and photos.
I'm still not satisfied, but photoprism seems to move into the right direction here. Digikam os great of you want everything on a single machine. Shot well has other advantages. None of them have a good solution to immediately and automatically import any photo taken on your phone.
Use apps like PhotoSync and it will upload automatically when photos are taken.
Re: Show HN: We built an end-to-end encrypted alternative to Google Photos
#418Earlier quoted context omitted.
They actually don't "sell".
They take your data and turn it into something that has value to them. With actual selling, that something is money. In this case it could be something else, but saying it that way will not help the general discourse of this problem at all. Much like being pedantic over terminology.
Ouch. This post reminds me of that one about GoogleSpeak: how Google limits thought about antitrust
https://zyppy.com/googlespeak/
Re: Show HN: We built an end-to-end encrypted alternative to Google Photos
#419what are currently the best open source projects that allow you to fully automate and manage deployment of your own personal (or multi-user) cloud photos/drive storage service? I found: [1]: https://github.com/nextcloud [2]: https://github.com/Piwigo with S3 extension: https://piwigo.org/ext/extension_view.php?eid=691 [3]: others https://arstechnica.com/gadgets/2021/06/the-big-alternatives-to-google-photos-showdown/…
Add https://lomorage.com, self hosted, cross platform, mobile friendly, support multiple accounts, and login from multiple devices.
Re: Show HN: We built an end-to-end encrypted alternative to Google Photos
#420Please please support custom storage back ends, I'd love to use my Dropbox or S3 or whatever to still fully own my pictures. And I'd love to pay extra to opt out of and analysis, tagging, etc of my photos. Basically I'd like the interface to be similar to Google Photos but with a privacy focused storage engine and clients.
you may want to take at look at: https://www.boxcryptor.com/en/