[AIT] Birdnet-go: Self-hosted realtime soundscape analyser for birds, bats and other wildlife
(github.com)
from chaospatterns@lemmy.world to selfhosted@lemmy.world on 10 Aug 05:44
https://lemmy.world/post/50488034
from chaospatterns@lemmy.world to selfhosted@lemmy.world on 10 Aug 05:44
https://lemmy.world/post/50488034
Birdnet-go is a real-time sound analyzer that detects bird species. It can take in audio from microphones. I set mine up to stream audio from my security cameras that I already had installed. There’s even a Home Assistant App, if you’re using it.
I live in a more suburban/semi-urban environment, but was really surprised to so many different bird species.
threaded - newest
OMG, birdnet-go is amazing. I had setup birdnet-pi, till an update to either docker or birdnet-pi broke it.
Now using birdnet-go and the stats are insane, this is just the last week <img alt="" src="https://lemmy.horwood.cloud/pictrs/image/309771bd-7a42-4d11-adcf-604e086b1734.png">
Nice. What kind of microphone setup do you have? I’m just starting with security cameras because I have those, but the audio quality is ehh.
Like you, using my security cameras. Riolink seems to have the loudest audio of the 2 I use. The other is a Unifi UVC G3 I had from work
My mom loves this so I build some additional microphones using github.com/Sukecz/ESP32-birdnet-mic and by 3d printing some casing for it to keep bugs out.
Someone share this with the bird voyeurs.
Stated Jake Farm
How does this compare to merlin? I dont currently have any reason to distrust the cornell university department of ornithology… but if they ever turn evil its good to know there’s a backup
Yeah, Merlin has been my go-to forever.
It is the same back end, but this one runs continuously on a server instwad of on your phone.
I’m afraid Merlin’s addition of videos of automatically running when birds are detected is the beginning of enshitification. Phone charge no longer lasts a full field day with this additional drain on battery. Bird of the day, x number of people listening, I don’t need of want any of this shit. Also dislike can no longer immediately cancel recording without saving, have to wait while application buffers or something before can exit without saving, for a long recording this can take more than 15 seconds. Pity, app is generally very helpful.
Ditch the AI slop logo, please.
Xale, the logo is pretty, on topic and quite cute. Is it made with AI? No idea. Does it really matter?
Give open source projects a break will you?
If you don’t like the logo just say so.
It is AI. 2min doodle would be better than this
I still like it and i have seen projects with much uglier logos made by humans. Nothing against humans or AI either.
An ugly logo made by a human wasn’t made by stealing from artists and disproportionately contributing to a climate crisis.
eta a qualifier I had assumed would be unnecessary…
Unless you want to argue that simply being alive is contributing to the climate crisis, I sure hope you don’t eat meat.
Or drive a car
You seriously don’t see the difference between a person existing and data centres pumping out slop made with stolen training days for the profit of billionaires? I’ll take a meat eating truck driver over an LLM defending vegan jogger any day.
I’m not defending LLMs, and neither are vegans. I’m pointing out that hating on datacenters for climate reasons if you then also eat meat is insane and detached from reality.
So you have to make zero pollution in order to criticize a big polluter? So vegans who drive cars should also keep their mouths shut?
No, but the meat industry next to datacenters is like a skyscraber next to a detached single family suburban home. It’s just so irrelevant. For all the reasons that datacenters are bad, pick one that doesn’t make you hypocrite.
Yes.
The device that youre using to type your message used materials that the world as well as slave labour and the internet you use is based on arrays of data centres that consume energy from who knows what energy source, so would you kindly keep your mouth shut?
<img alt="https://drkt.eu/files/yet-you-participate.png" src="https://drkt.eu/files/yet-you-participate.png">
That’s literally what you did. People voiced their concern over data centres and you brought up the meat industry.
I understand that reading comprehension is difficult but I do wish you best of luck!
And now you throw insults. Great job. You’ve completely dismantled my argument.
Would you prefer I brought up:
You can’t just argue against one issue and brush aside the others. From my point of view, use of generative AI borders on indefensible.
Yes, I would prefer you brought all of those up instead! I am not pro-datacenter- I just want you to pick an issue that doesn’t make you a hypocrite.
They did express their displeasure. Even said please
That’s true…
That’s literally what I did.
I guess if you spend enough time chatting with sparkling autocomplete you forget how real conversations work.
🔗 AI Logo
🔗 and probably everything else too
🔗 Including the readme
*Takes a look at the first three folders*
.agents/skills.claude/skills.cursor/rulesYep…
The maintainer of the project probably won’t see it here, but that repo does appear to be fairly active, they’ve got thousands of merged PRs. If you contribute a better logo they might replace the AI one.
It’s an AI program specifically to use AI models to recognise birds. If you’re against AI you should probably learn to recognise them manually.
“AI” is a pretty bad blanket term. BirdNet / Merlin is a model trained specifically on bird calls (and some noises) built to recognize bird calls. App and logo use huge general-purpose models to create content of questionable quality.
And yes, a decent birder should never trust just Merlin ID.
Yup… HUGE difference between human curated machine learning and generative AI. AI sure is a buzzword these days, which is why it gets used in marketing, but I think doing so is confusing people who don’t know enough to distinguish between the different flavors of “AI”.
Oh you’ll smoke a turd in hell for your insolence. That would make sense. It’s much easier to complain.
<img alt="" src="https://lemmy.ml/pictrs/image/12718aab-ee65-43f7-8985-c1e987e2a3e4.jpeg">
Totally agree, I threw a quick filter on some public domain, actual human artwork, and cropped it square. Took me as long as it would have taken to generate that slop logo.
Here’s the original source: Citation
For the love of anything even mildly good or wholesome…enough of the fucking AI slop images!
Rule 7+8
Let this post be the one to ask a question that’s been bothering me (not a coder):
This and similar claims in other software suggest - to me - that the AI is running locally on the raspi, no need to exchange data with sites providing AI APIs or whatever.
I struggle to believe that this is possible, let alone on a raspi.
So, how “local” ar the AIs that help more and more software with various tasks?
Is “local AI” some sort of selling point that is often false or deliberately misleading? “Localwashing”?
edit: thanks for all the answers, this has been enlightening.
I haven’t verified it, but it local ML models on audio is pretty efficient and should run fine on a raspberry pi.
If you’re speaking of audio analysis ML or AI models, they are quite efficient. As a mediocre musician, who is clinically deaf and cannot hear some frequencies with enough clarity to make a judgement either way, I lean on AI assisted audio mastering. I haven’t dabbled in AI generated music, but with the right DAW and a handful of plugins, some keyboards and instruments, you can do a lot right in the comfort of your home.
It can run on device on your phone birdnet.cornell.edu
Listen, everything that they are doing by scorching the planet and turning the tap water brown, could be achieved on CPU or much smaller GPUs on local machines - it’s a question of focused, tight training. What they’re doing instead is literally attempting to build gigantic models that take in absolutely everything (burning real books in the process because fuck history we got money to make) and attempting to sell the idea that this brute force, one-thing-does-it-all approach, which requires us to give up everything else that is good in the world, is the only way forward, and yes, we must build the roko basilisk now.
It’s a big, big grift.
Didn’t downvote you however, are you saying if I ran an AI model local, that scorching the planet and turning the tap water brown is ok with you, just because it’s local?. Because that’s exactly what we’re are doing when we access the internet. Throw AI out of the scope, and we’re still pumping massive amounts of carbon in to the air, still turning the water brown. There are 8.4 billion people on this planet. Approximately 6.3 billion of them are online. That’s just individual humans. That doesn’t include the hundreds of thousands of data centers that aren’t even connected with AI and all the energy production to make them run.
So, who will be the first to actually practice what you preach and shut down your servers and internet devices? C’mon…make the first step towards a cleaner planet.
This whole internet could run on a tiny fraction of what it current consumes, with no drop in service levels… if the extraction of profit from every single interaction was not prioritized.
That being said, I was addressing someone who refused to believe that local inference was even possible on CPU; this tells me that the efforts to present the current grift as the only and inevitable AI outcome is working, on minds that do not fundamentally grasp what the technology is, nor what it’s doing. I think the tech is great, and highly useful - the way they are implementing has perverse incentives and existential outcomes, and it does not have to be that way, but it is, because capitalism.
But it doesn’t.
But it is.
So, now what are we going to do? Give me real world solutions that will actually fix the issue, which would be implemented by real world corporations and real world data centers. I’m not problem hampered, I’m solution based. So let’s fix it.
Decouple people from big tech, allow people to run, fix, and maintain their own hardware instead of replacing it on an industrial scale every 6-9 months and centralizing ownership in data centers.
My entire stack runs off a 14kw solar array with a battery. Net negative. It’s totally doable.
Edit: also I tested my stack against a stimulated 1k active users and couldn’t get it to break a sweat, so I’m not saying everyone needs to do things this way, but your techiest uncle could host the entire family in a privacy cooperative, for instance
I’m not making fun of what you’ve said, and I’m not talking down to you. That said, I like where you’re going, but that will never happen. You can mark it on the calendar: ‘That irmadlad asshole said - never happen’. On the one hand I applaud you loud and clear. On the other hand, we live in a global consumerist society. Corporate is just giving them their fix. That’s their plug.
I have solar too. Hell I have solar on two barns. I like it. I often think about how far along we might be here in America if a lowly peanut farmer had not been ridiculed beyond belief for putting solar on the white house. Now, they were not what you’d call efficient, but it was a start. I think the next presidency took them down.
Oh no doubt. I’ve long since been a promoter of the notion of a server in every home. We live in a digital society. We have digital assets. These need to be centrally housed on a very secure server and backed up.
However, self hosters and homelabbers make up <2% of the world’s population. It gets even less if you differentiate between consistent homelab/self hosters and those who were just experimenting. Yeah, we do it, but not on a massive scale to even make a ding. Again, I like where your head’s at, but it’s not us we need to convince, it’s them. And they like their crack. Corporate likes their money.
I mean, I’m not saying give up. Keep pushing. But realistically, we’re in a script.
Saying it’ll never happen and giving up won’t make that 2% get any bigger
Make the tools easy enough that the irritation of the original platform gives them enough of a push to try something easy to use.
If you want to help beta test, I’m working on an appliance-level automated installer for provisioning home cloud servers.
Got friends who need help? Help them or send them my way (or to your nearest techie)
Again, you get a standing ovation from me for your drive. Emphatically, I hope you and others with your drive succeed. I’m just giving my perspective of being involved with people and their computers since the mid 70s. I’m no IT guy by any stretch of the imagination, but I like helping others with their digital devices, because it is my desire for everyone to use said digital devices in the most safe, private, and anonymous manner possible. If any morsel of my experience can help someone, then I’m all about it.
However, and for instance: It’s been back at the beginning of the year, I had almost a 2 hour ‘training session’ with a friend on how to effectively use UBlock Origin. Something so very simple but quite effective. But you have to ‘train’ it to get the most out of UBlock Origin. Three weeks later I’m over at his house and he’s disabled UBlock Origin because it was annoying. Work, yes it takes a little work to learn how to drive the bus, and that’s where most people fall by the wayside for just about anything, not just digital devices.
I’d be willing to take a look at it. I’ve got a couple empty boxes around here. Back when Microsoft released their Home Server edition, I thought ‘Finally! I hope this catches on’. But they killed it. I’ve always thought some bright, young, entrepreneur could take a couple of small ThinkCenters, construct a server environment, put it in a box, with almost a fool proof, one touch button that went out on your network, connected all devices to that one small central server, deploy apps etc, and everything would be jippity jippity. The problem, however, with making things fool proof is that fools are quite ingenious little buggers.
That’s my specialty! My goal with my projects has always been to lower the barrier to entry while compromising as little as possible on things that matter. I remember everything Google and Apple built before they started dumbing it down and removing features. I remember how much effort went into onboarding users and making sure people could adopt things easily. It didn’t always work, but I’ve been very successful in making my entire de-google’d stack useful to my own family.
The hardest part is done, now I just need to make it replicable for others to spin up their own. If you’re interested, I’ve got a repo scaffolded but I haven’t published the actual stack installer yet: git.circuitforge.tech/Circuit-Forge/…/issues
I’m not trying to throw shade on your project nor am I trying to discourage you in any way. I think it’s a noble ambition. I hope you make a mint doing it. If I live long enough, I’ll be able to tell people ‘Yeah, I met the dude on Lemmy before he was so successful.’ But if 71 years of experience has taught me anything, it’s that it has to be so easy a cave man could do it - as if it just fell in your lap, and it has to be work and trouble free. Otherwise, the average Joe Schmoe citizen isn’t going to spend the time. Having a secure network takes work. Keeping up with updates and making sure they aren’t going to make the wheels fall off, takes work. Work is where the vast amount of people fall off the wagon.
I bookmarked your project and I’ll give it a serious look see. I will tell you that the demo site: oasis.circuitforge.tech, returns a bad gateway Error code 502 from Cloudflare.
Rock on with yo’ bad self.
It’s a small model that runs on your machine using a TFLite model. AI/ML models come in many different sizes. Some more conducive for running locally than others. There’s more info on the detection pipeline here.
I didn’t write this, just found and self hosted it.
Thanks.
I understand now that this software does not use one of the big general AI providers but a specialised model (for birds esp.?) that runs locally; that I can get behind.
Are there other keywords in software descriptions I can look for that help me make such distinctions?
Specialized models can be made incredibly efficient. The big tech guys are trying to replace humans which is why their crap takes such huge servers to run. Building local-first/only just means you need to specialize the model and the pipeline to match what you’re doing
Thanks.
I understand now that this software does not use one of the big general AI providers but a specialised model (for birds esp.?) that runs locally; that I can get behind.
Are there other keywords in software descriptions I can look for that help me make such distinctions?
LLM and VLM are generally the type of model that needs a ton of resources to run. Birdnet uses an audio classification model, but there are dozens of types of models.
I’m building specialized task-focused pipelines which keep the memory footprint small for self-hosted users, and I’m working on a unified framework too make building other pipelines and applications much easier: docs.circuitforge.tech
If you give me some details on what you’re after I might be able to find you some projects or at least clarify keywords
Edit: an example of how efficient these can get, I have a little ESP32-driven waveshare display running an ONNX wake-word detection model which is absolutely nuts. 4mb psram and a dual-core 240mhz cpu and its only job is handling “hey Mycroft” and establishing a proper speech-to-text connection with my server downstairs
Audio processing by ai can absolutely happen on a pi.
As an example, whisper.cpp is a high performance speech recognition model, and it does a great job. The biggest version of it as a model requires less than 5GB to live fully in memory, the large model at about 1.5billion parameters, and it works great even CPU only.
The one used most often is small, which has about 244million parameters, and needs about a gig of ram and nothing more.
For a pi 5, usually the base or tiny model are used. Small could be used in most scenarios, but tiny and base are able to run real-time.
So it depends for how local - sometimes its all self contained and can do just fine on a pi, sometimes its local in that it needs an LLM endpoint to hit, but that can be running on another system entirely.
It would really depend on the project for how/where/why for a definition of local.
Thanks.
I understand now that this software does not use one of the big general AI providers but a specialised model (for birds esp.?) that runs locally; that I can get behind.
Are there other keywords in software descriptions I can look for that help me make such distinctions?
Aside from “local model”, you might see things like whisper.cpp, llama.cpp, ollama, or “OpenAI compatible endpoint”, or a reference to getting models from HuggingFace (you can think of it like github for llm’s - its more and less, but accurate enough for what we’re talking about here).
whisper.cpp is a c++ implementation of whisper (ASR as mentioned), llama.cpp is a high performance and lightweight tool to support llm/llm inference, ollama is a very user friendly way to do the same, and OpenAI compatible endpoint is just using the protocol for communication, so you can point the client at llama.cpp or similar running on the same machine or another you have access to.
Most things with an extemely tight purpose like this (like listening for some birds singing) will use much leaner models, and may be able to run even from a phone.
I’ve been running birdnetpi on a raspberry pi 3 for years. It runs very well on this modest hardware. I also run birdnet-go in a docker container and use my reolink security cameras for the RSTP audio feed to that instance.
There are all sorts of AI/ML models available that can run on smaller hardware. But they are usually very specific and not general purpose, like the LLMs.
There is also this similar software, which is a little clunky but not slop github.com/Nachtzuster/BirdNET-Pi
For the benefit of others:
A quick general awareness comment:
A post about a project not by the original creator is not a promo post. A user will not necessarily know the details to provide a disclosure unless a project explicitly has on in their repo.
Requiring it on a project post like this (“Here’s a cool thing I found!”) is not doable.
That said, its clearly an AI project start to finish - the detection wouldn’t really work without it. So I think [AIT] is appropriate here.
@chaospatterns@lemmy.world can you please add the [AIT] tag?
Edit: Thanks!
If you need something on the go I quite enjoy whoBIRD (available from F-Droid). Wonder if these use the same model.
I use both Song Sleuth and Smart Bird ID (iOS). Fairly decent. I wouldn’t bet any significant bet on it’s accuracy 100% of the time. I’m no birder, but it seems to work well enough for my purposes.
.