It seems to me two longstanding Wishlist items would greatly reduce the need for person detection for people like myself, thus saving the drain of recurring costs for both Wyze and many customers:
I believe that both of these could be implemented on the camera chip firmware as a one time investment by Wyze to strengthen weakened customer loyalty and strengthen the long term finances of the company.
The cloud processing power needed to analyze the clip to see if it contains a person.
Regulard motion & sound detection are done in the camera I believe.
@jls4wheeler, you’re welcome!
@toddreg, regular motion detection is just, “Yes, something moved. Record and upload the video. Now store it.” I don’t have all the details about the expense differences but part of it is, “Yes, something moved. Record and upload the video. The video has been uploaded and stored. Run the algorithm on the video. Is there a person? Yes > Send person notification. No > Send regular notification.” There’s more activity happening and it’s more complex.
@sodcam, I’ll bring up those Wishlist items. We have been discussing them internally. Thanks! 
Thank you for the email and the open-ness regarding the balance of features and implementation cost, as well as being up front about how you are trying to creatively solve the balance.
As a suggestion, perhaps you would also consider an “opt-out” choice for Person Detection. That is an option whereby a user sets a camera to never want Person Detection for that particular camera. How that would be done, and perhaps how that would be rescinded are obvious questions. I for one do not need Person Detection on any of my three cameras based one how they are deployed, and don’t force the need in the future.
Obviously no one can predict how many people might take advantage of an opt-out, and maybe it doesn’t even make a significant dent in the cloud workload required, but just a thought.
My questions
I have to say I’m kinda disappointed. I still think Wyze is a great company and will buy products from them. I will probably support Wyze each month with some money. I just don’t like the looks of where this is headed.
Make the RTSP firmware better, so that your users wouldn’t NEED to use the Cloud based Person Detection at all!
I personally think anyone would should be ashamed to sue Wyze over this. The cameras only cost $20. Get real!! I challenge anyone to price out person detection on any of the others. I am happy with what we got. My only complaint is trying to view prerecorded video from a Wyze Cam without taking the SD card out is not real easy. That actually works better on Arlo, though the Wyze cams are better for pretty much everything else
Agreed - this is an obvious solution in addition to reducing the cloud footprint significantly for those that use it. Saavy users can add any feature they want post processing in their own infrastructure.
This is why Wyze should put it on the camera itself therefore eliminating the cloud costs
I’m with 8043c9b245a0d18ab05c, give us some idea of $ amount that allows for us to carry our weight of the expense. And I appreciate your approach. A bit of an amateur mistake to promise $0 before knowing the actual cost. But out of all the home automation companies, you’re the only one that would :
A. Promise something for free (before knowing the cost),
B. Admit the mistake, but still offer to keep the promise,
C. Not just slap everyone with a cost + profit fee, and justify it with your low prices on
all the equipment.
So thanks. I for one am glad you got a finance guy. Just don’t give him too much power.
I agree! Wyze is still a great company and they are in a thought position. I wish they would have gotten a finance guy before they started free person detection. Maybe he would have said you won’t be able to afford it for free.
Our finance guy is super nice and he’s on board with our dreams of making smart tech accessible for everyone. He’s also much better at giving those doses of financial reality as we need them. 
He’s a great addition for the team and we are well-balanced.
Remember the MaxDrive? yeah neither does Wyze.
They seem to have problems with letting you keep your data local
We were really excited about MaxDrive. We were also super sad when that partnership fell through. 
They can’t/don’t know how, at least, in the small amount of memory/cpu they have.
XNor.ai did it, which what was they were bought out.
Too bad Wyze did not have a better contract/lawyer with that.
Personally, I think they should have an external local hub they sell that can do the processing “local”.
(Probably what will happen with their outdoor cams)
The cams can report their events to the local hub, the local hub can process the image/video, and then report to the user if it detected something.
I would love to go the RTSP firmware if that firmware was up to par with the non-RTSP firmware. just too many issues with RTSP firmware to use it reliably yet.
Any chance Wyze will try this again? That would be a solution… Person Detection on MaxDrive. Camera sends video to drive. Drive says person or not. This would make notifications come over faster and no cloud costs.
If they brought back the project of Maxdrive…
I can’t make a promise on this but I know we’ve kept looking for a good partner for it. We haven’t given up on the concept. Though I’m not sure if we’d be able to use it for person detection.