Camect - New “do all” box

NJP

n3wb
Joined
Dec 28, 2014
Messages
2
Reaction score
3
Any thoughts on this new box? Allegedly allows local storage for all the big cloud cameras (arlo, dropcam, etc) as well as most other IP came. Includes 1tb expandable storage as well as event detection (by type). Looks promising. The one reason I got rid of my cloud cams and went to a closed network with Blue Iris was privacy and monthly cost. This might negate that.

 
Last edited:

aristobrat

IPCT Contributor
Joined
Dec 5, 2016
Messages
2,982
Reaction score
3,180
Allegedly allows local storage for all the big cloud cameras (arlo, dropcam, etc)
From their FAQ (snip below), doesn't sound like their ability to record big cloud cams locally is official or should be depended upon.
Nest and Arlo cameras work too, on an experimental basis. These services could someday make changes that prevent them from working with Camect
Frequently Asked Questions - camect
 

mat200

IPCT Contributor
Joined
Jan 17, 2017
Messages
13,649
Reaction score
22,741
Any thoughts on this new box? Allegedly allows local storage for all the big cloud cameras (arlo, dropcam, etc) as well as most other IP came. Includes 1tb expandable storage as well as event detection (by type). Looks promising. The one reason I got rid of my cloud cams and went to a closed network with Blue Iris was privacy and monthly cost. This might negate that.

Hi @NJP

Cloud cameras, unless they're using open standards that you can use locally ( rtsp ONVIF etc.. ) I would not consider them as stable platforms for on-premise recording as I have experienced numerous service outages from big name cloud providers as well as my ISP.
 

looney2ns

IPCT Contributor
Joined
Sep 25, 2016
Messages
15,521
Reaction score
22,657
Location
Evansville, In. USA
Any thoughts on this new box? Allegedly allows local storage for all the big cloud cameras (arlo, dropcam, etc) as well as most other IP came. Includes 1tb expandable storage as well as event detection (by type). Looks promising. The one reason I got rid of my cloud cams and went to a closed network with Blue Iris was privacy and monthly cost. This might negate that.

Meh. They are still after a monthly fee.
 

TL1096r

IPCT Contributor
Joined
Jan 28, 2017
Messages
1,223
Reaction score
465
Everyone is after monthly fees. Many software like photoshop charges monthly fees. I was surprised when BI was a 1 time payment. Good guy not charging monthly/early when he could have easily.
 

spectator

n3wb
Joined
Oct 19, 2017
Messages
25
Reaction score
23
Everyone is after monthly fees. Many software like photoshop charges monthly fees. I was surprised when BI was a 1 time payment. Good guy not charging monthly/early when he could have easily.
FD: I'm one of the Camect founders.

Camect's subscription isn't required. You get a perfectly usable system with functionality comparable to BI (but targeted at a less sophisticated user) with no subscription. BI does give you alert delivery, whereas we don't ... but on the other hand we give you real local object-detection-based AI and BI does not.

If you get object detection from Sentry, and automatic software updates from BI, I believe you'll have to pay more per year than you do with Camect's $60/yr subscription (but that was not the point of our fee). You'll also be sending your data to the cloud for the Sentry AI, whereas ours is 100% local.

Our monthly fee covers things that depend on internet infrastructure that is an ongoing cost for us -- things like WebRTC-based remote access, alert delivery (email, Telegram IM, or app notification), backup to services like Dropbox/Google Drive, etc. We include these services for two cameras with the purchase of the box, but if we don't charge for them, they aren't going to be sustainable long-term as we're selling the hardware as cheaply as we can -- and we're not in the business of making money by selling our users' privacy.

Re: Integrations with Nest and Arlo ... We want people to be aware that (like anything else with a cloud camera) the manufacturer can cripple them at any time but we'll keep them going as long as technically feasible. We are serious about keeping them going though.

The point of supporting these cameras is to give people stuck on Nest/Arlo/Ring a way to see what they can get from Camect, and encourage them to switch to regular IP cameras eventually because they'll get better quality, lower cost, and no dependency on the cloud while keeping the same user experience.

The point of Camect as a whole is to provide an experience comparable to (and hopefully better than, at least as far as the AI is concerned) Nest/Arlo/Ring for users who currently have no other choice. Setting up a PC and configuring BI is quite beyond the technical reach of many people out there.
 

NJP

n3wb
Joined
Dec 28, 2014
Messages
2
Reaction score
3
FD: I'm one of the Camect founders.

Camect's subscription isn't required. You get a perfectly usable system with functionality comparable to BI (but targeted at a less sophisticated user) with no subscription. BI does give you alert delivery, whereas we don't ... but on the other hand we give you real local object-detection-based AI and BI does not.

If you get object detection from Sentry, and automatic software updates from BI, I believe you'll have to pay more per year than you do with Camect's $60/yr subscription (but that was not the point of our fee). You'll also be sending your data to the cloud for the Sentry AI, whereas ours is 100% local.

Our monthly fee covers things that depend on internet infrastructure that is an ongoing cost for us -- things like WebRTC-based remote access, alert delivery (email, Telegram IM, or app notification), backup to services like Dropbox/Google Drive, etc. We include these services for two cameras with the purchase of the box, but if we don't charge for them, they aren't going to be sustainable long-term as we're selling the hardware as cheaply as we can -- and we're not in the business of making money by selling our users' privacy.

Re: Integrations with Nest and Arlo ... We want people to be aware that (like anything else with a cloud camera) the manufacturer can cripple them at any time but we'll keep them going as long as technically feasible. We are serious about keeping them going though.

The point of supporting these cameras is to give people stuck on Nest/Arlo/Ring a way to see what they can get from Camect, and encourage them to switch to regular IP cameras eventually because they'll get better quality, lower cost, and no dependency on the cloud while keeping the same user experience.

The point of Camect as a whole is to provide an experience comparable to (and hopefully better than, at least as far as the AI is concerned) Nest/Arlo/Ring for users who currently have no other choice. Setting up a PC and configuring BI is quite beyond the technical reach of many people out there.
Thanks for the reply. I have been following your releases for awhile, and glad to see you post here. Camect *probably* isn’t targeted at the demographic typically frequenting IPC, and it sounds like you know that demographic and have a solid business model/plan.

I wish you luck, and wish you were around five years ago when I got started!!
 

fenderman

Staff member
Joined
Mar 9, 2014
Messages
36,897
Reaction score
21,250
FD: I'm one of the Camect founders.

Camect's subscription isn't required. You get a perfectly usable system with functionality comparable to BI (but targeted at a less sophisticated user) with no subscription. BI does give you alert delivery, whereas we don't ... but on the other hand we give you real local object-detection-based AI and BI does not.

If you get object detection from Sentry, and automatic software updates from BI, I believe you'll have to pay more per year than you do with Camect's $60/yr subscription (but that was not the point of our fee). You'll also be sending your data to the cloud for the Sentry AI, whereas ours is 100% local.

Our monthly fee covers things that depend on internet infrastructure that is an ongoing cost for us -- things like WebRTC-based remote access, alert delivery (email, Telegram IM, or app notification), backup to services like Dropbox/Google Drive, etc. We include these services for two cameras with the purchase of the box, but if we don't charge for them, they aren't going to be sustainable long-term as we're selling the hardware as cheaply as we can -- and we're not in the business of making money by selling our users' privacy.

Re: Integrations with Nest and Arlo ... We want people to be aware that (like anything else with a cloud camera) the manufacturer can cripple them at any time but we'll keep them going as long as technically feasible. We are serious about keeping them going though.

The point of supporting these cameras is to give people stuck on Nest/Arlo/Ring a way to see what they can get from Camect, and encourage them to switch to regular IP cameras eventually because they'll get better quality, lower cost, and no dependency on the cloud while keeping the same user experience.

The point of Camect as a whole is to provide an experience comparable to (and hopefully better than, at least as far as the AI is concerned) Nest/Arlo/Ring for users who currently have no other choice. Setting up a PC and configuring BI is quite beyond the technical reach of many people out there.
This is a very good concept and a low price point. I would suggest replacing the vstarcam on your recommended list - that junk and its clones dont belong in any home. As far as the accuracy of the analytics, there is no way to know at this point as it is pre-sale. Ultimately over the next few years the analytics will be camera side . IPVM just had a favorable write up on Wyze and dahua Analytics+ functions Dahua Analytics+ Tested
 

spectator

n3wb
Joined
Oct 19, 2017
Messages
25
Reaction score
23
This is a very good concept and a low price point. I would suggest replacing the vstarcam on your recommended list - that junk and its clones dont belong in any home. As far as the accuracy of the analytics, there is no way to know at this point as it is pre-sale. Ultimately over the next few years the analytics will be camera side . IPVM just had a favorable write up on Wyze and dahua Analytics+ functions Dahua Analytics+ Tested
Yeah, I mostly agree with you about the Vstarcam. That's a pretty old recommendation and at the time we added it it was quite cheap compared to the others. Now that we support Wyze it might be a better "cheap camera to play with" recommendation. The only downside (for an unsophisticated user) is that for the Vstarcam it's literally plug and play (you plug it into the network, our box finds it and configures it -- and we do this for a number of other brands too) ... With Wyze, you still have to use their app to get it onto the network before we can do anything with it.

Re: Analytics ... We're not perfect (and no one is), but we think we're better than Nest/Arlo/Ring and one difference is that we actually allow you to submit examples when things don't work -- people who put up with a few rounds of that will usually see an improvement in their specific problems.

As far as "no way to know [about analytics] since it's pre-sale" .... This is a fair point, and we've been thinking about how to convince people that it's not vaporware. If you have a camera you'd be willing to make accessible over the net, I'm willing to hook up a box to it and send you the alerts so you can see what you think. (I can do this for up to about 6 cameras if anyone else is interested.) For nest users this is really easy (and less risky) because you can just share a camera with us and we can use it.

If you do this, please do keep in mind that this is a beta product and still improving - and that for the best results it's helpful if you'll put up with a few rounds of training - but the system is real enough that we're not afraid to show it to people.
 

TL1096r

IPCT Contributor
Joined
Jan 28, 2017
Messages
1,223
Reaction score
465
FD: I'm one of the Camect founders.

Camect's subscription isn't required. You get a perfectly usable system with functionality comparable to BI (but targeted at a less sophisticated user) with no subscription. BI does give you alert delivery, whereas we don't ... but on the other hand we give you real local object-detection-based AI and BI does not.

If you get object detection from Sentry, and automatic software updates from BI, I believe you'll have to pay more per year than you do with Camect's $60/yr subscription (but that was not the point of our fee). You'll also be sending your data to the cloud for the Sentry AI, whereas ours is 100% local.

Our monthly fee covers things that depend on internet infrastructure that is an ongoing cost for us -- things like WebRTC-based remote access, alert delivery (email, Telegram IM, or app notification), backup to services like Dropbox/Google Drive, etc. We include these services for two cameras with the purchase of the box, but if we don't charge for them, they aren't going to be sustainable long-term as we're selling the hardware as cheaply as we can -- and we're not in the business of making money by selling our users' privacy.

Re: Integrations with Nest and Arlo ... We want people to be aware that (like anything else with a cloud camera) the manufacturer can cripple them at any time but we'll keep them going as long as technically feasible. We are serious about keeping them going though.

The point of supporting these cameras is to give people stuck on Nest/Arlo/Ring a way to see what they can get from Camect, and encourage them to switch to regular IP cameras eventually because they'll get better quality, lower cost, and no dependency on the cloud while keeping the same user experience.

The point of Camect as a whole is to provide an experience comparable to (and hopefully better than, at least as far as the AI is concerned) Nest/Arlo/Ring for users who currently have no other choice. Setting up a PC and configuring BI is quite beyond the technical reach of many people out there.
Thanks for the reply! I will be honest that I just saw a monthly fee and didn't look into it much so thank you for taking the time to explain.

Can this box replace or be an alternative to a VPN? I know some that would never run a VPN and use ARLO. If this box will make them more secure and still keep all the features they want like 2-way audio I will purchase. And with the Camect you would not use the arlo app or cloud based recordings once it is setup?

Thanks
 

spectator

n3wb
Joined
Oct 19, 2017
Messages
25
Reaction score
23
Camect does provide an alternative to a VPN, but it cannot make an Arlo or Nest more secure -- because your cloud camera is still going to send its data to the cloud. Camect works by fetching the data back from the cloud -- These integrations are mainly a way for people with Nest or Arlo to try the Camect feature set without immediately having to buy new cameras -- but to be cloud-free, they do eventually have to move to IP cameras.

For people using regular IP cameras, Camect does provide an easy to set up (i.e. no set up required at all) alternative to VPN that should be just as secure -- assuming you trust our software (just as you have to trust the vpn software) and you set a local password that would be comparable to the password for your vpn. We're arguably a little less reliable than a VPN because in order for webRTC to work, our servers (which do the first authentication check and help the two endpoints find each other) have to be up, whereas a VPN is completely independent of anyone else's infrastructure. However, for unsophisticated users for whom a VPN is too much hassle -- both to set up and to use -- this should be a big improvement over any other way that they could access their IP cameras, such as port forwarding or P2P over some shady Chinese backend.

webRTC also works in modern web browsers, and we support chrome and safari ... so we provide a web-like experience without having your data in the cloud. (Basically, your web server runs a javascript app which then sets up a connection directly with your box at home to get to your data. Video is never accessible to our servers, and latency is usually significantly better than going via the cloud, although the connection takes a bit longer to set up than when you're connecting to a server in the cloud.)

We don't support two-way audio yet. For now, we record audio and video. Someone using our device with an Arlo would typically use our UI to view video or alerts but would still have to use the manufacturer's app for two way audio or any feature that isn't a "camera" feature (e.g. to know that the doorbell button was pressed on a doorbell). If the person with an Arlo later got some IP cameras, another nice thing is that they'd get to see everything in the same UI with the same feature set, making it that much easier to get people to gradually convert.
 

looney2ns

IPCT Contributor
Joined
Sep 25, 2016
Messages
15,521
Reaction score
22,657
Location
Evansville, In. USA
Yeah, I mostly agree with you about the Vstarcam. That's a pretty old recommendation and at the time we added it it was quite cheap compared to the others. Now that we support Wyze it might be a better "cheap camera to play with" recommendation. The only downside (for an unsophisticated user) is that for the Vstarcam it's literally plug and play (you plug it into the network, our box finds it and configures it -- and we do this for a number of other brands too) ... With Wyze, you still have to use their app to get it onto the network before we can do anything with it.

Re: Analytics ... We're not perfect (and no one is), but we think we're better than Nest/Arlo/Ring and one difference is that we actually allow you to submit examples when things don't work -- people who put up with a few rounds of that will usually see an improvement in their specific problems.

As far as "no way to know [about analytics] since it's pre-sale" .... This is a fair point, and we've been thinking about how to convince people that it's not vaporware. If you have a camera you'd be willing to make accessible over the net, I'm willing to hook up a box to it and send you the alerts so you can see what you think. (I can do this for up to about 6 cameras if anyone else is interested.) For nest users this is really easy (and less risky) because you can just share a camera with us and we can use it.

If you do this, please do keep in mind that this is a beta product and still improving - and that for the best results it's helpful if you'll put up with a few rounds of training - but the system is real enough that we're not afraid to show it to people.
I'd be willing to give your product a trial run over time, but I'm not up to offering up a camera to port forward.
See my other reviews, to see if I'm worthy. ;)
 

spectator

n3wb
Joined
Oct 19, 2017
Messages
25
Reaction score
23
Do you have to port forward with this?

I am trying to get my friend to remove arlo and buy this with the box for more security:
https://www.amazon.com/gp/product/B01M0IHF9K/ref=ox_sc_saved_title_9?smid=AM1AKWRN957PC
Just in case there was any confusion ... Port forwarding is not something you'd ever need to do using Camect the way it's supposed to be used -- i.e. recording from cameras on your own network. It was mentioned purely in the context of wondering if anyone wants to try the analytics with us remotely before we're shipping boxes. In that scenario we have the problem of the box and the camera not being on the same network.
 
As an Amazon Associate IPCamTalk earns from qualifying purchases.

TL1096r

IPCT Contributor
Joined
Jan 28, 2017
Messages
1,223
Reaction score
465
Thanks. I think this is great for beginners too that can switch to something intermediate and want to use this instead of BI. Good job if it all works out and I will be a customer for my friends with less secure cams.
 

SAVideoman

Getting the hang of it
Joined
May 20, 2016
Messages
45
Reaction score
51
Re: Analytics ... We're not perfect (and no one is), but we think we're better than Nest/Arlo/Ring and one difference is that we actually allow you to submit examples when things don't work -- people who put up with a few rounds of that will usually see an improvement in their specific problems.
Sounds interesting. Are you using a GPU to do the AI analytics processing?
 

spectator

n3wb
Joined
Oct 19, 2017
Messages
25
Reaction score
23
Sounds interesting. Are you using a GPU to do the AI analytics processing?
The AI analytics are mostly on on the CPU at the moment. We do use the integrated intel GPU extensively for encoding and decoding though. No support for an external GPU yet.
 

SAVideoman

Getting the hang of it
Joined
May 20, 2016
Messages
45
Reaction score
51
The AI analytics are mostly on on the CPU at the moment. We do use the integrated intel GPU extensively for encoding and decoding though. No support for an external GPU yet.
Thanks spectator.
Are you using something like YOLO for you AI? That algorithm seems pretty popular lately.
 

spectator

n3wb
Joined
Oct 19, 2017
Messages
25
Reaction score
23
We're using SSD + MobileNet with quite a bit of our own tweaks ... When we started working on this, the results were much better with SSD (given the constraints of our platform) than YOLO v2. Since then YOLO v3 came out and we need to redo the comparison at some point, although the SSD results are pretty reasonable, so we haven't put a lot of priority on doing that.
 
Top