Path: csiph.com!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail From: rbowman Newsgroups: comp.os.linux.misc Subject: Re: Hmm ... Ordered PI-Zero 2WH "Kit" - Possible Problem Date: 28 May 2026 17:40:19 GMT Lines: 24 Message-ID: References: <4rsfemxvi6.ln2@Telcontar.valinor> <7XCdnUfd6c-ExYv3nZ2dnZfqnPWdnZ2d@giganews.com> <10v6cj6$2l317$1@dont-email.me> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-Trace: individual.net 2qsQ6WBan6YVg4fw/rdK/Qvmr5uY9p2XrFVvm5FxgLjGoFqEm2 Cancel-Lock: sha1:vRXDmQb68IzBBJeGeXUIcqC7XHA= sha256:H9sB9FQCT+0BUXh7Y8yeUbEq4mt5ZoKYVXF9Dc5jvKY= User-Agent: Pan/0.162 (Pokrosvk) Xref: csiph.com comp.os.linux.misc:87238 On Thu, 28 May 2026 03:44:26 -0400, c186282 wrote: > 'Sensors' aren't a problem anymore. They can be very tiny and very > efficient. Most of the spectrum and electric/magnetic too. > > But how to INTEGRATE that info ? You can 'see' frequencies with a waterfall display and a SDR to scan but that couldn't be transferred to a continuous awareness. It's interesting that processing 'wake words' like 'hey Alexa' do a similar translation and convert the spoken word into a spectrogram. Neural networks are very good at image classification. After the device recognizes the wake word then it can phone home for more computing power to handle the rest of the conversation. I don't know if AI is involved but I've been playing with a free phone app called 'Seek'. It's similar to the Google photo recognition but specialized for recognizing flora and fauna in your geographic area. Take a photo and it comes back with 'Wilcox's penstemmon' or 'Large Flower Clarkia'. It also correctly identified felix cattus. It's handy since I'm involved in a project to weed out non-native species from an area. I know the usual culprits but when it gets to grasses like cheatgrass I'm a little vague.