Please consider making this work without the cloud (but still accessible from the cloud if wanted of course).

You say you are selling camera hardware, so could selling a RPi or Google Coral board with pre-trained models be feasible? Nicely packaged up with a nice case etc - people don't need to know it is a RPi in a box etc. Store images/video locally with optional "cloud backup" as an paid-for add on?

I have had basically everything in my house shutdown before when my ISP had a "maintenance event" - could not turn lights on, could not use a baby monitor, could not turn the heating on etc etc because everything wanted to talk to the cloud even though my lights and heating are physical things inside my house.

Apart from that, some nice online integration would be good - IFTTT, MQTT (bonus points for local broker support to avoid cloud), and a public API etc so people can wire it up to their home if they want (e.g. unrecognised face at door? => turn on lights, dog on lawn? => turn on sprinklers etc etc)

Are you suggesting not using cloud for privacy concerns? Based on the feedback from people's comments here, I realize we should do more to alleviate the privacy concerns. Curious on to know your thoughts about the following aspect: As I mentioned, we only store short video clips corresponding to events that the user created (we already discard the other motion clips that are deemed as irrelevant by our models.) We also allow users to delete the the alerts they have received and when they delete each alert, we permanently delete the corresponding video clips from our dbs... would that alleviate your privacy concerns?

We actually built our first prototype using RPi, we tried 3-4 different RPi cams, the image quality of all of them was very poor. Also the final cost would much higher than the cameras we are using right now...

Supporting IFTTT is in our near term road map. Appreciate the suggestions!

It was more that if my internet connection goes down, are the cameras useless? What if the internet/AWS is just "slow" one day - will the notifications be delayed significantly making any "reactive" integrations pointless/ludicrously delayed? If things can run locally (doing inference for multiple cameras via a single "box" you plug in to your WiFi router etc) then you can be super-fast with IFTTT integrations.

My main line if thought was that I built basically your product for spotting when a cat climbed into my plant pots using ML and a RPi3 - the idea was that when it saw the cat, it would squirt a water pistol at it to scare it away - inference on the RPi 3 was too slow (if I was doing this now I'd use a coral accelerator maybe) and by the time it realised a cat had got into the plant pots, the cat had already taken a shit and left. I worry that your product might suffer from similar end to end latency. Niche use-case? Perhaps. I have Amazon Blink cameras here and the IFTTT integration is delayed by about 30 seconds so by the time you get a notification there is someone at your door it is to late to do anything as they will.have already left/kicked the door in by then etc. Doing all this locally would be super fast

My main concern was not really about privacy - you'll need to cover GRPR if someone from the EU happens to walk into frame of one of your customers' cameras one day in the future anyway (Good luck)

Love that use case :) re your point about latency, that is one of the main reasons we are doing all the inference on the cloud. Almost all deep learning models (at least CV models) need GPU to run with low latency. That's true that if you have set up an automated response from another device, it may still work if the internet is down but for alerting the user, you would still need internet connection even with a central hub...

FTR, a simple RPi4 yields ~4FPS on a DL object detection model. You may need cloud for your central hub, but neural network inference can be done locally.

By 4FPS do you mean 250ms latency per frame? What OD model specifically did you get that latency from on RPi4?

It was my understanding that GP's main concern was the service shutting down should his internet fail, and not one of privacy.

I see. That is actually a great point and we have heard that concern from others as well. The cameras we are offering now come with local storage. We have been thinking of adding some capabilities when camera is offline but since we do the inference on the cloud, the smart alert and video indexing features (our main value props) would not work offline.

Than what is your selling point against all those Chinese cloud based services?

Our main value prop is more powerful event detection capabilities that what anyone else is offering (and at a lower price.)

There will be many kinds of customers, but among them those who value privacy and those who want convenience. One of them will drive the bulk of sales, the other maybe not. On the other hand, if this is a value prop that excites people you’ll surely have the incumbents consider the economic threat and may add that feature as needed.

Ya I think finding the right balance between privacy related risks and cost/convenience/features is a an important aspect of this space.

Not OP, but if you want privacy conscious users to use your business, they will want to be able to host it themselves.

Ya I see your point. We have tried to take the privacy seriously from the beginning (and now realize we need to do more) but as far as not using the cloud at all, that would be a significant limitation on the type of inference that can be done locally and I think the benefits may outweigh the associated risks for many users/use cases...

I have a number of cloud enabled cameras now, but I’m also building my own system to keep everything local-only.

If you want to sell me an ML-based system, you’re welcome to train the models in the cloud, but they have to run on local-only assets. And you have to give me complete control over downloading new models periodically to a machine of my choice, and the updating my local devices.

Thanks for the feedback. What are you using the cameras for?

Even storing short video clips in the cloud is not good enough.

If you want people like me to buy it, you've got to cut the cloud out of the picture. Give me something that does inference on the device. Sure, give me an easy way to send selected videos back to you for training data, and feel free to push optional updates to the camera with updated models, but I'm not giving you a raw video feed of my home or business.

The reason we went with the cloud is not just to use the data for training data. 1- we wanted to be hardware agnostic so our solution works with any hardware without any dependency on hardware specifications. 2- Doing deep learning on the device would require GPU which would significantly add to the initial cost for consumers. 3- as I mentioned in my response to another comment, regardless of where the inference is done, the recorded clips will have to be stored on the cloud so if a bad actor comes in and take the camera with them, the user can still access the recorded video clips after the fact to know what happened. Even though we do inference in the cloud, we do not store the video clips unless they correspond to an event that the user is interested in--we discard the other clips. So I'm not clear on how doing inference on the device would have any advantage from a privacy stand point.

I get it. Moving things to the cloud keeps the devices cheap and keeps them from becoming out of date. It also makes you identical to other large players and makes me think you don't have much of an advantage. Make the devices expensive. Put privacy first. It might just pay off with higher sales. Your alternative is racing to the bottom in a competition with Amazon, Ring, etc. You will lose.

Sure, you can push the video to the cloud, but encrypt it on the device. Let me control my data.

Based on the feedback from people's comments here, I realize we should do more to alleviate the privacy concerns. Curious on to know your thoughts about the following aspect: As I mentioned, we only store short video clips corresponding to events that the user created (we already discard the other motion clips that are deemed as irrelevant by our models.) We also allow users to delete the the alerts they have received and when they delete each alert, we permanently delete the corresponding video clips from our dbs... does that help alleviate your concerns?

Not really. Having a device send video frames to the cloud basically kills the deal for me. If you did something to the video first... maybe. Let's say you train a model to process video. You come up with a network where the first few layers are fixed and they perform a non-reversible transform of the video into some sort of symbolic representation. I might sign up for that. I want to know that my video/audio feed isn't being used by a 3rd party, and that it isn't being used by you for something other than the intended purpose(i.e. to model my behavior to sell me things, to create profiles of my activity, etc). I understand that my first proposal doesn't actually address all of those concerns, but I am willing to trade privacy for convenience to a very limited degree.

I'm basically working on something like this for myself. I have a Nvidia Jetson Nano that I'm trying to train to tell me when my garage door is open without my wife or I present, when my laundry is done, and whether or not the lights are on.

That is pretty interesting but non-trivial. We'll have to think about it. Thanks for the suggestion. I should mention our business model is not based on monetizing users data. I'm interested to know more about your personal project. Feel free to shoot me an email if you are interested in having offline discussions and bouncing ideas off each other: rafiee at visualone dot tech

Just because that’s your model today, that doesn’t mean you won’t get bought by another company whose sole purpose is to use all assets for gathering as much private data as possible and using it for their own purposes.

If you want to sell to people who actually care about privacy, then you have to build your systems so that it is impossible to abuse them or use them in any other way. Or, at the very least, it is extremely difficult to abuse them.

That is a valid point. We will think more deeply about these. Thanks for the feedback.

I just want to say that it's good that you're listening to people here, but at the end of the day you have to go with what's pragmatic for your business.

Not to say you shouldn't move forward with your grand plans, but there are lot of people, myself included, who won't touch any of this for those reasons.

I don't need more of my life stored on someone elses network.

I hear you and understand the concern and appreciate the feedback. What do you use the cameras for if you don't mind me asking?

Seems very similar to Camio (https://camio.com/) but targeted at the consumer space. I like the clever wrapping of object recognition and motion detection into real world actions- "dog getting on couch". I think this can do well in the consumer space.

It's hard to bring this technology to enterprise because there are already large companies offering similar video analytics capabilities combined with complete video management solutions. See Axis, Briefcam, Milestone, Gorilla, Agent Vi. And a lot of the VMS first companies have integrations with a spread of analytics companies. You really don't want to spend your dev time creating yet another VMS. Work on being able to integrate easily with others.

One huge advantage I see with you guys is that you can do everything in the cloud. Most of the industry analytics companies require dedicated hardware or servers with GPUs. But, these companies are targeting large installations- 100s-1000s of cameras across multiple sites. Your technology might work great for 1-2 residential cameras, but I'm not sure how much it can scale to industry.

Best of luck, anyways! I like the concept.

That is a great analysis. I completely agree with everything you said. We are currently targeting consumer and SMB use cases and not enterprise/large scale applications at the moment. Thank you!

> Our solution can also alleviate privacy concerns since we only store short video clips on the cloud for alerts corresponding to user’s events of interest instead of for every motion detected.

Do better: allow the owner to store everything locally instead of in the cloud.

Well the problem with that is many people use the cameras for security also. So if a bad actor comes and takes the camera with them/breaks it, there would be no way for the users to get access to the footage after the fact to know what happened.

The video wouldn't be stored on the camera, but an NVR or something similar. But "locally" could also include a cloud storage solution of the end user's choosing in addition to actually on prem.

If someone manages to steal one of my cameras I'll have some close-up, HD video of them stored on my hard drive.

I see. So you mean for the VMS solutions, not the stand alone IP cams (such as Nest, Ring, Wyze, doorbells, etc)?

I don't know what @nkrisc, but I definitely mean: camera should be able to store data on any standard sftp server. Or on a CIFS server for the Windows dudes. Or WebDAV.

Want to store it in some proprietary cloud? That's fine, but it should only be done as an afterthought. SFTP, CIFS, and WebDAV would enable any power user to build and use their own home storage or cloud storage.

I see the value of what you are proposing but that would be useful only for tech savvy users (vs mainstream)...

If you also sell the optional local storage device (a network video recorder), then at least you give your customers the option.

Otherwise, you are not materially different from Ring or Google.

Thanks for the feedback. Our cameras have local storage but we do the inference on the cloud. Our main value prop is more powerful event detection features beyond what Nest is offering (Ring is not really offering any as of now.) and at a significantly lower price...

I 100% disagree. Mainstream people ask tech savvy people to help all the time. And, importantly, paid-for cloud services should IMO also make their services available over sftp. Not doing so is a large contributing reason I don't use those services.

Congrats on launching. Some really fascinating ideas. Although I suspect my biggest concern would be, the consumer electronics space is not fun to be in.

I'm curious if you've considered targeting small business at all, that maybe had enough flexibility that consumers could adopt. I'm a board member of my condo corporation (HOA), and our staff can struggle quite alot with scrubbing through security camera footage to find events, mainly damage to the property. So having a system that can say load all the video of a vehicle passing through a garage door for the past two days would save a ton of time. Of the feature you've described, on when something goes missing or damage appears to the build would save lots of time.

Similar to alot of the sentiment in this post, because its monitoring of public spaces, I likely wouldn't accept a cloud based solution. It's just difficult to vet that companies are doing the right things and have a good security posture, despite privacy policies. Simple cloud storage like an S3 bucket would probably be fine.

Anyways, best of luck on the launch.

Thanks for the feedback. We have been looking into some SMB applications including property management. Would love to know more details about your use case if you are open to an offline discussion. We are actually storing on S3 (and only store short clips for user's events of interest in the cloud.) Thank you!

Sure, my keybase contact info is in my profile. I'm also @kevin in the Kubernetes and CNCF slack workspaces.

Great. I will follow up with you. Thanks!

There are existing commercial solutions that can address the HoA use case you describe here. Lots of smarts in the local NVR device on the network, or in the admin station that connects to the NVR device — follow the person in the yellow jacket, ALPR, and so on. Then you can easily add off-site storage, or whatever.

The problem is that these commercial solutions are expensive.

You could try to build your own with OpenCV and other tools, but that’s a really hard task if you’re not also selling for the same kind of money.

An example commercial solution would be https://www.youtube.com/watch?v=s0BLx9eKAyk . Powerful and expensive. Most businesses do not want video uploaded to the cloud, but some are receptive to it, and the large business analytic vendors are developing products for the cloud.

Thanks for sharing.

It's an interesting space, it strikes me immediately as the video version of Minut[1] I got one their first-gen units and was pretty disappointed, but the principle has a lot of potential. What I have really wanted to see from any of these IoT devices is a workflow that goes - tell app you are going to teach it to look for something (or listen), stage an occurrence, confirm the device registered, then stage it again, and expect a notification. So for example, "this is what its sounds like when the clothes dryer is done" - dryer plays a little song - app says 'got it' - dryer plays its little song - notification is delivered saying "your dryer says it's done". I don't think that's beyond the capability of the hw and sw, yet I haven't seen a good implementation yet...

1: https://www.minut.com/product/features/

That was actually my vision initially when I started working on the problem. I wanted to allow users to train their cameras for things they care about. But later decided to start with events that we can detect without requiring a lot of input from the users initially...

I can see advantages to both, I am sure plenty of users want plug-and-play and would be annoyed by the idea of having to work through manually training. Maybe there can be an "advanced" or "developer" mode... Even just having a feed of events that I could go into and add tags or other info to might work.

I like the idea of having a different tier for people who are more tech savvy/patient. We currently allow users to provide feedback on alerts (thumbs up/down) which we use to make the models more accurate over time but training from scratch would require a lot more input from users...

I've been waiting for something like this (and daydreamed about doing something similar as a hobby) - what's holding me back from ordering immediately:

- How do I hook up an outdoor camera? Mounting instructions? Does it need power? Wifi based? (guessing it will be: "straightfoward instructions", yes, yes, but would still like details)

- A privacy statement at the very least - and ideally privacy from the ground up - perhaps via differential privacy, or maybe you allow users to pay less if they make their unencrypted photos available to your training models. IMHO privacy concerns are what are really holding back smart home tech and keeps me from adopting it.

Thanks for sharing. Valid points and very interesting! We have been mostly focused on building the technology before the demo and have been deferring tasks like adding more details to the website, finalizing privacy policy, etc. I expected that the focus on HN would be mostly on the technical aspects of the product, but I am realizing that HN actually cares most about the privacy aspects. We will make these a higher priority going forward.

To answer your first question, the installation/requirements for our cameras are pretty much the same as the other stand alone security cams in the market (Nest, Ring, Wyze, etc.) Basically, they require power (plug into outlets) and require wifi connection.

Smart of you to target Nest cameras, since those customers don't care about privacy as much as HN readers that comment on threads like these do.

There is plenty of academic research done on this stuff (I know, because a group at my department did these things for elderly care). Have you looked into this, or are you making your entirely own thing from scratch?

Thanks :) we are definitely trying not to re-invent the wheel. We have been using various open source tools that are available and have been building on top of those... if you know of something that you think would be particularly useful for us to look into, I would love to know about it.

Dipak Surie was at my department. His work focused, as I understood it, on learning what tasks people were trying to perform and what actions were included in those. Say that a person with dementia may be trying to make coffee, and would forget what they were doing halfway through. The system would understand that and help out by saying what the next action should be.

https://scholar.google.se/citations?user=-mxqfbIAAAAJ&hl=sv&... is his Google Scholar page.

Thanks for sharing. Sounds pretty cool. We will look into it!

This will be major as a home assistant integration. Right now the nest camera component is broken due to the recent Nest=>Google migration. This has broken streaming nest cam footage for everyone that migrated. I'm actually surprised a 3rd party has actually gotten support for nest cam stream API before home-assistant itself.

Anyways, when it comes to security cams in the consumer space you have

Low: Wyze, Yi, other Chinese Mfgs Mid: Ring High: Nest

But I think the major money is in the prosumer space: The custom Dahua (Lorex/Flir) or anything with an NVR. Massive false positive rate, bad detection algos, shitty software. Have you seen Lorex's Rapid Recap feature? Definitely cool but too expensive. Also Dahua NVR's have had security concerns too

Rapid Recap: https://www.youtube.com/watch?v=ZIWnQ9arSJg CCTV Hacked: https://www.youtube.com/watch?v=A9ea9fllnME

Smart local secure powerful NVR replacements is what business + prosumer home owners are looking for, good thing is that most IP cameras are running on a handfull of standards, instead of a seperate API for each brand.

How did you get access to the NEST API streams?

Thanks for sharing. I had not seen those. Need to look into them more closely. We have been focused on stand alone cams. Nest shut down their APIs last year but they allow users to share their camera streams.

So you require the Nest streams to be public then???

yes but a random url that you get is private (only people you share it with would have access) and you can change/invalidate it anytime.

All leading camera brands have on SoC motion detection (Honeywell, Motorola...) and they have ML driven detection on their on premise stream servers. This is going to be a very tough sell.

There are various cameras on the market claiming to have smart alert features but in reality the only AI features that they are currently offering (that we are aware of) are person detection, unfamiliar face detection, facial recognition, activity zones, some limited object detection... if you know of any solutions offering beyond those, we would love to know about them... also these features come at a high price both in terms of the initial hardware cost and monthly fees (e.g. Nest Cam IQ selling for $300-$400)

Axis has super advanced AI/ML video stuff. I've seen demos of it doing things like "use a network of PTZ cameras to auto track the person who left an unattended bag".

They have lots of free applications in their App Gallery, and with minimal effort cheap Chinese cameras can be made to work with the system (since Axis OEMs their cameras from them anyway). https://www.axis.com/en-us/products/camera-applications/appl...

Thanks! We will look into it.

One super annoying thing I've noticed about OpenCV - no ability to read relevant metadata from streams. It seems like lots of applications of OpenCV would benefit from using the motion capture functionality of these cameras (for example, streamed over RTSP via OnVIF) as a pre-screening technique, but I've had trouble getting access to it.

It seems like it should be straightforward to add, but I haven't had a chance to do so.

what metadata specifically are you trying to get from the streams?

The OnVIF specification allows data on motion to be transmitted with frames as metadata. This is really useful for, for example, IP cameras.

I'm in the industry and I have to agree here. All of the devices we develop and use (third-party) either already have this capability built-in to the device or are working on it right now.

I agree it is an active area and other people are working on the problem, but can you be more specific about the capabilities that you have seen in other products?

Essentially all of the events you mention plus a good amount more are either in an active product development pipeline for vendors we work with or are actively being worked on internally on devices we are building.

It's a tough space.. margins on these hardware devices are thin already, and sometimes negative as they are simply loss leaders hooking in users for long-term saas payments, so in our case, it's cheaper to build it in-house.

Good luck, however!

I completely agree with you re margin on the hardware being low. We are not planning on making any money on the hardware. But I also believe it is not possible for startups/companies that don't have copious resources (like Google and Amazon) to do great at both hardware and AI (i.e. build cutting edge AI/CV tech)... Thank you :) If you'd be up for discussing this further, may I ask you to email me? (rafiee at visualone.tech)

I see huge opportunities for this type of technology to be used for Security operations at Houses of Worship, Sports Arenas, and Large Buildings if you're thinking of a B2B angle.

FWIW, I've seen a few other companies pop up offering a similar service to that market and are doing well. Lots of security operations centers are still manually run w/ 100s if not 1000s of cameras being monitored by a team of humans (to the best of their ability).

Good luck! This is an awesome idea!

Thanks. Appreciate the feedback and the kind words! We will look into the use cases you mentioned.

Great work. If you can make the service GDPR compliant and get certification you are probably 2 steps ahead on any big brand where it concerns Europeans. Distrust towards built-in AI on devices like Ring and Nest is growing. Is offering a white label version on your roadmap?

Thank you! Great feedback. We are actually GDPR compliant but don't have the certification yet. The cameras we are offering are indeed white label cameras. They are the same cameras as the ones sold by Wyze...

Congratulations on launching. It is a very fascinating tech space.

Could you elaborate how the system as a whole is compliant with GDPR and other European privacy laws?

I ask because I explored and eventually decided not launch a computer vision product (in 2014) due to compliance aspects.

Looking at it from the whole system perspective, with a camera pointed at your neighbour's garden or the street in front of your home it is quite difficult to make a compliant system:

First, there are the hurdles of the GDPR (it has to be a legitimate purpose, the subject has rights and must be informed etc....).

Second, there are the broader privacy laws for public spaces, where it is mostly illegal with exceptions for government and some specific use cases for banks etc. (I am familiar with Danish rules, not those of every EU member state).

I did meet a startup some years ago that claimed that their computer vision was not video surveillance since they did the video stream processing on-device and only emitted events (not video) to the network, so perhaps there is a way to do it nowadays.

I would love to hear your perspective on the current compliance concerns for this type of computer vision systems.

Thank you! I guess I should not have said that as we have not looked into all the nuances yet (some of which you mentioned), but we have taken many extra steps to design the product/architecture with the main principles in mind from the beginning. For example, what we store, how we store them, allowing users to have control of their data (as an example, as I mentioned in another response, we allow users to delete the alerts and when they delete an alert, we permanently delete the corresponding video clip), and of course if a users decides to delete their account, we permanently delete all of their data.

Then I'd expect at least a link to a privacy policy somewhere on the site?

You're absolutely right. We have been finalizing the details. we will add to the website asap. If you have additional specific concerns/thoughts, I would definitely like to know what they are.

My main question would be: where do you get your training data? A large source of unease with ML tech on inputs from users homes is based on that and the processing around it.

We are using some publicly available data and have labeled some data ourselves. We are not using users data for training currently.

Why don’t you partner with Wyze, Nest and Ring instead of trying to go direct to consumer with your own camera? Direct customer acquisition and operational complexity of dealing with hardware, inventory, etc. seems outside your core business. Wyze would seem like a great first start since they lost their person detection partner Xnor to an Apple acquisition some months ago and they have millions of cameras in the wild today. Really all of them are trying to build the type of detection you have.

That is a great suggestion and is indeed our plan even though we are also offering our cameras now :) Thanks!

If you have any hopes of selling your software to fortune 500 enterprise class customers, it needs to be software, not software as a service, able to operate self-contained, as in capable of running on an air gapped network without any Internet connection. It should also integrate with Genetec, the VMS industry leader.

I agree with that. We have not been looking into enterprise applications much yet. The main limitation of using self-contained software is it would be dependent on hardware specifications (especially GPU) and can pose significant limitations and a lot of complexity.

I have a smart home startup running in the middle east, and we want to develop the exact same idea in the future. The only difference is the cloud part. People here (and I assume many other countries) hate the cloud, they want everything running locally and with a minimum subscription. So we've been brainstorming a solution similar to camect[1] that handles everything locally. We still have a long time to reach that since the market here is at least 5 years late that USA.

Do you consider building this kind of device in the future?

[1] https://www.indiegogo.com/projects/camect-world-s-smartest-m...

Thanks for sharing. Yes we are aware of them. I’d like to know more about what you’re doing. Please shoot me an email if you’d like to talk: rafiee at visualone dot tech

This is exactly something I have been looking for a long time. At least I thought so, until I got to the "cloud" mention.

You missed the train with cloud. People are becoming increasingly privacy concious, and more and more are adpoting "once its on the net, I have lost control of it".

I fully understand that getting the clips to teach the neural network is important from your side, but this is dealbreaker even for people that would otherwise be willing to pay monthly just to use the software locally. Perhaps asking for "send us this clip please, so we can train the software" from time to time would be acceptable.

If your concerns is being hardware agnostic, consider selling the "local only" option with strong disclaimer that only narrow subset of hardware is supported.

I see your point and appreciate the feedback. The main issue with local only is having it to work reliably on different types of hardware would be a major undertaking itself (aside from the technical restrictions that come with it) and would not allow us to focus on building the AI...

can you put up an "about" page on the site? I would like to know a bit about the team and the founders before I purchase home cameras from them.

Absolutely. We have been heads down working mostly on the product to get it out before the demo day but we will make that a high priority. thanks for the feedback!

One of the things I noticed that you're offering is cloud storage for the last 7 days. If you're building a consumer facing product (vs. trying to have your technology/company integrated into one like Nest) I'd recommend exploring additional options here. Storage is so cheap and it always bums me out that the offerings are so bad or expensive. For example, I feel as though I should be able to hook up my blink, ring, etc. all into the same S3 bucket for my own long term storage.

Also - I don't see a privacy policy on your site. Are you using people's video to further enhance your models?

I agree storage should not be as expensive. I think many companies use that as a leverage to make users pay a subscription fee (due to lack of services that’d compel users to subscribe.) We are finalizing our privacy policy and will add ASAP. We are not using users data for trading models for other users currently but we do use their feedback on the alerts to improve the models for their own events.

Seeing a lot of companies doing these for b2b applications. Are you fully focused on b2c?

We are looking into various areas and are trying to find the best niche to focus on initially--does not have to be b2c necessarily. do you have any specific b2b application in mind?

I think its a decent although not particularly unique idea. How much of a grasp do you have on the actual tech though? Detecting things like dog chewing on a a shoe is not trivial, just putting it through a model you found and edited a bit on github probably isn't going to cut it. Also you already have these AI powerhouses with strong technology who are capable of making something similar and haven't found that much success (e.g. nest)

I agree the idea is not unique at a high level but this is an unsolved and difficult problem. We are building a solution block by block. The only event detection features Nest is offering is person detection, unfamiliar face detection, facial recognition.

right, so you guys are mostly trying to stand out with the tech then, that's totally cool. Good luck and all the best!

Thank you!

This is something I've been thinking about doing for a while. How soon after the event occurs does the user get a notification?

Our end-to-end latency is about 3-4 seconds now but we know we can bring that down to less than 2 seconds in the near future and maybe even less than 1 second in longer term.

Best of luck Mohammad Rafiee! This is really neat. Is there a reason you guys are pricing so low?

No "About" page, no privacy policy. Maybe I'm being too cynical, but there's probably a lot more to be made off collecting data from the cameras than you could expect to charge consumers.

The lack of an About page is weird.. I don't think you are being too cynical, the lack of information about this company on their site is concerning. Especially considering they are selling cameras you put in your house.

I'm also surprised by the name they chose, as Visual One is already a software product by Agilysys. I would be surprised if it's not already copyrighted?

Thanks for the feedback. We have made it a high priority to add more details to our website asap.

You're absolutely right. We have been mostly focused on the technology side and have been trying to move fast and learn from our users. I agree we need to invest more in providing more details especially on our website. Appreciate your feedback.

Re your point about user data, we are not planning on monetizing on users data--we currently only store 10-second clips corresponding to the alerts for each users events of interest (not even every detected motion) when they happen so the users can view later.

Thanks :) we need to figure out the right pricing but we did not want to lose some potential users as a result of pricing too high. what do you think would be a reasonable price?

Hey! You did a great job! I also think it's too cheap. As you are dealing with security, people need to be sure it will work so maybe higher price would be better. Do you have competitors? I would advise you to be around the same price.

Thank you :) That is a great point. Yes in the consumer space the monthly subscription fees are mostly in the $5-$10/month range and that's why we decided to go with $7 but for SMB applications we will set the price higher (with more features.)

> We currently support Nest Cams and also offer our own cameras (same as the cameras sold by Wyze) with indoor and outdoor options.

How are they your cameras when they are Wyze's? I think you may want to correct the language here to eliminate any confusion.

That is a great feedback. I will change the wording in the post to clarify but they are actually white label cameras that Wyze is selling and we are using the same white label cameras.

Can I hook up my existing Wyze cams or install your firmware on them?

Technically that is possible but it is a very involved process (and does not always work), also you would not be able to use Wyze's service anymore...

Thanks... you've probably got lots on your plate right now will keep an eye on Visual One! If there is enough demand I'm sure you'll invest more in the firmware retrofit, or maybe I'll buy some of your cameras

Just to confirm, it can count when it’s a new person but if they enter again it will know it’s the same person and not count them again as new ?

That is correct with the caveat that their face must be visible/detected. At the moment, we count the number of unique persons using facial recognition.

Can you make an add-on temperature sensor that detects and alerts for fevers? This is what we need urgently, everywhere

That is a very interesting use case with everything that is happening right now. We are exclusively focused on computer vision but I agree that would be very valuable for a situation like this.

Unfortunately, you have to make your Nest streams public to be able to use it. I’m unwilling to do so.

please note that the url you get is random and only people you share it with would have access. Also you can change/invalidate it anytime.

Envysion is the b2b version for restaurants/retail if anyone is curious.

Thanks for sharing. We will look into what they are doing.

can the model be trained for something more complex like shoplifting?

We have heard that specific use case quite a bit. Detecting shoplifting is very hard as there is not much visual distinction between a customer picking up an item to buy and someone shoplifting, at the moment they pick up the item. It would require tracking the person and verifying that they paid for the item which can have a lot of complexities...

Great job, you are addressing a huge problem! What about B2B?

Thank you! We are now exploring consumer and SMB but may also look into B2B use cases more closely at some point. If you know of any good b2b use case, please do let us know.

Is this ONVIF compatible?

Not yet but that is in our roadmap. Please feel free to email me at rafiee at visualone.tech and I will keep you posted as we add the support.

Thanks for sharing. I need to look more closely but on a brief look it seems the only event detections they are offering is person detection, car detection and people count...

Substantive, thoughtful critique is welcome on HN, and many users have been posting such comments in this thread. That's great.

This comment, though, breaks the site guidelines, and that's not ok. Please read and follow them when posting here: https://news.ycombinator.com/newsguidelines.html.

We detached this subthread from https://news.ycombinator.com/item?id=22560528.

p.s. In addition, could you please stop creating accounts for every few comments you post? We ban accounts that do that, which is also in the site guidelines. You needn't use your real name, but for HN to be a community, users need some identity for others to relate to. Otherwise we may as well have no usernames and no community, and that would be a different kind of forum. https://hn.algolia.com/?sort=byDate&dateRange=all&type=comme...