Lifetime Premium Member
- Aug 19, 2017
- Reaction score
OP wins best "middle finger" post of the entire year on RDF.
This is an interesting point...I’m still saying that a lot of these BSM’s could be filtered on a superhet if new programming logic And software were developed. Not to the effect of Thea but still... it could be done.
Definitely would have made it a lot better, but probably would have reached the limits of the hardware too quickly.This is an interesting point...
Would it be possible to take a "legacy detector" with an FPGA, train an AI on BSMs and real radar guns using the RF front end, and run a NN on the FPGA to classify the signals and filter out BSMs?
Obviously the RF front end in Theia is much more sophisticated than anything we've seen before, but is a single conversion receiver with I/Q demodulation really required to filter out BSM? Or is it overkill for that application?
To put it another way, if Escort had hired Rob and Sterling, could they have saved the Max series?
I agree with you. A lot of BSMs could be filtered out without AI. It would take longer of course, but it's definitely doable.Thanks for sharing Jon.
I’m still saying that a lot of these BSM’s could be filtered on a superhet if new programming logic And software were developed. Not to the effect of Thea but still... it could be done.
Right, and where your implementation has a distinct advantage is it's very manageable. You're able to respond to changing RF emitters very quickly while others must develop a means to cope via more traditional methods.I agree with you. A lot of BSMs could be filtered out without AI. It would take longer of course, but it's definitely doable.
Also interested in this. It's not a Bushnell, but it is another basic Part 15 gun. Does the signal become "cleaner" when complexity is added to a gun?
When a gun is overdriving the receiver distortion appears as what looks like sidebands (spikes to the left and the right of the main carrier). It kind of looks like a trident.How close was the gun to the receiver? Are there oscillations that may exist and be detectable when gun is really close to detector?
Escorts (Max's) have been said to have a problem with the Bushnell when it's very close to the detector, or is it likely an over driving issue if anything? I've even noticed my Uniden and NRDSP being finicky if the Bushnell is triggered very close to the detector. I've never really noticed it with my Redline though.
Genesis.....no different from a door opener or a Bushnell to a legacy detector.
That's the idea man. We have been kept in the dark so long about this stuff that we have been forced to take manufacturer's word for so many things that were wrong. Well, that didn't work out too great for users like me who desired freedom from BSMs and a detector with better range.
Bingo I'll smoke to that. Hell I'll even splurge and sip on a good whiskey.Also, I just want to reiterate for the Escort execs viewing this thread - I have patents pending on use of machine learning and AI in radar detectors, quadrature/complex front ends in radar detectors (6 different ways of using complex data), and CW discrimination by name presented to the UI and for filtering purposes. And I will defend them aggressively.
Vortex I don't know if seeing that is an option, but I heard them say, you will be see/display as little or as much as you want...Wow. Yeah I remember Rob mentioning you guys are able to see handshake and things like that. Does that affect Theia's ability to recognize the gun as a Stalker?
In practice when out driving, do you think you'd be able to still see that? I'm sure there's a lot more going driving down the road vs sitting in a lab. Do realworld dynamics and noise impact your ability to still correctly inference these guns?