Yeah, and I suspect you might also agree that not crippling (via subscription requirement) Edge AI Person Detection on Cam v4 would be following another tenet:
Operating Value: Maximize Every Component
If the Edge AI is in the device, then according to my read of the situation it’s doing Smart Focus for non-subscribers, but that’s about it, and it seems like it could be doing more.
These really are good things to keep in mind, I think, and I hope that some Forum users (in addition to Wyze employees) take the time to read them and understand what this company is saying it’s supposed to be, as well.
Wyze long ago abandoned their mission statement they just haven’t bothered to remove it. But like the original web site quotes about the V4 they will probably just delete it without comment.
Bottom line is any on device feature is irrelevant to state once you make it a feature only available with a cloud based service.
That’s where they effed up. Who cares what the on device feature is if you have to have cloud services in order to use it? 30mp? Available in future cloud service packages only. That’s the logic here.
YES! I have often overlooked that value because it can be a complex and subjective one about what counts as maximizing the component when some things need balance. For example, they had a discussion back with the V2’s and V3’s about how they didn’t have the resources to include both RTSP and all their AI/Cloud stuff effectively on the same camera because each took up too much of the processor/ram/storage to be able to do both. So which way is “Maximizing” the component in a case like that…it is totally subjective to the individual. There are also considerations about things like overclocking a processor in a computer…in some ways overclocking is “Maximizing” the component, but someone else might think that setting clocking limits MAXIMIIZES the component’s longevity and reliability, etc. So I often gloss over that value because of the contradictions and complexities involved in deciding what that means, especially in relation to the other values when 1 might not mesh with the other. However, in this case, I think it’s pretty clear that this subject fits perfectly in that value. I’m glad you pointed it out, because I often skip over that one because it’s usually so easy to debate the other side.
As for smart focus, I don’t think that it is related to the Edge AI, though I wish it was. I think it is just done with the main area of motion detection. It would be awesome if it was limited to only person detections or Person/pet/vehicle. I have suggested that to them. I think the smart focus is mostly being limited to cameras that have sufficient processor/RAM and isn’t related to edge AI, at least that was the impression I got when they first launched it, but I could be wrong.
Edge AI. Faster notifications, more accurate detections, while supporting Smart Focus.
I guess your thinking probably depends upon how much you want to trust the accuracy of that statement.
Anyway, that’s why I made my previous statement about the relationship. Maybe Edge AI isn’t completely responsible for the Smart Focus feature, but Wyze is saying that it’s involved in some way. If it’s already there and doing something, then let’s maximize it!
I understand your perspective about the subjectivity of statements like that, too. There’s always going to be compromise somewhere, and it’s impossible to please all users all the time with whatever decisions they make, but they can still strive to be “too good to be true”.
Interesting, it does imply that it is related to Edge AI, but then why does it mostly zoom in on motion instead of just a person? Maybe they’ve updated it in the V4. When I tested it more on the V3Pro, I didn’t really like it because it wouldn’t necessarily recognize and follow a whole human form from my experience. Now I’m curious. Maybe it does do more than just motion…but if so, why would the smart focus be free and included for everyone and not paywalled if the Edge AI is paywalled? The smart focus still works even if they don’t allow the Edge AI. That seems like a contradiction. #AMA2ASK
Personally I find this value to be a little cringey in the wording because in colloquial use this phrase generally means that something ISN’T actually TRUE, and what Wyze really means is to make something that is so good it FEELS unbelievable but is actually true (amazing tech at an affordable price). While the phrase has roots that means something that sounds so amazing that it obviously isn’t true (get rich quick schemes/scams, Miracle Weight loss products, free vacation offers, etc). But I know what they mean, and their intent is what matters, so I’m okay with it and try to be supportive…I just get vibes of scratching nails on a chalkboard when I read that phrase.
I hadn’t thought of it in that way, but I see your point. Totally valid. At least they’re not abusing “literally” in the same phrase.
I described my experience with it (when I finally got it to work for the first time) in another topic:
I actually thought it was pretty cool when I saw it work on my own camera for the first time, but it’s not a feature I imagine I’ll use a lot.
I haven’t played with it a lot but in my limited experience it kinda seems to do both, just depending on what’s causing the most motion—maybe like what’s the largest blotch of changing pixels? The cameras without Edge AI are already doing motion detection with pixel differentiation, though, so I don’t know the answer to that or why Wyze would say that Edge AI is present on Cam v4 for “supporting Smart Focus.” I don’t have the answer. That’s a good question.
Sometimes when I load a Cam v4’s live view, I’ll see that Smart Focus has zoomed in on a portion of the view even when no person is there, usually just a blowing tree branch or something. Perhaps that’s why it’s still labeled as a “BETA” feature.
I don’t think that’s what they’re saying, or at least that’s not how I’ve read it. Elsewhere I’ve referred to how at least one Wyze Help Center article refers to “the full feature list of Edge AI”. Based on what I’ve read in Wyze’s communication and my own experience, I still think that when I enable Smart Focus I’m using one of the features on that Edge AI list, and I don’t read “Edge AI” as being equivalent to “Person Detection”; I think Person Detection is only part of what Edge AI is supposed to offer.
I don’t know the technical details, though. That’s just how I read it, and I think the answer to your question might be that only part of Edge AI requires a subscription (which seems consistent with what Wyze is apparently trying to say), but that’s just an educated guess. I think that’s another good question to ask.
These really are good things to keep in mind, I think, and I hope that some Forum users (in addition to Wyze employees) take the time to read them and understand what this company is saying it’s supposed to be, as well.
cool when I saw it work on my own camera for the first time, but it’s not a feature I imagine I’ll use a lot.