Use Sound and Machine Learning to Identify Birds with a Raspberry Pi - BirdNET-Pi

Поділитися
Вставка
  • Опубліковано 25 січ 2025

КОМЕНТАРІ • 49

  • @vanik0r0
    @vanik0r0 11 місяців тому +4

    Hi Tim, I've spent over a month trying to get various BirdNET-Pi installations working and to date, yours has been the only one that consistently worked...well-done!
    Can you do further instructions to work out how to change the password and harden-up the security? 😉

  • @pixelfrenzy
    @pixelfrenzy Рік тому +1

    What model was the "more omnidirectional USB microphone" that was mentioned in the 3D-printed case section? Alt question: what kinds of USB mic or soundcard are recommended? Presumably frequency response isn't much of an issue given the frequency range of birdsong? Thanks.

  • @mxcollin95
    @mxcollin95 2 роки тому +1

    Very cool idea and project!!! 🤙

  • @pantac4493
    @pantac4493 2 роки тому +3

    Great videos. Thank you

  • @davidverheyen6635
    @davidverheyen6635 Рік тому +1

    Thank you, this could be something for our nature-visitor-center. I understand that it is possible to use the microphone from a birdbox-cam rtsp-stream. Is it possible with this software to play the realtime audio from the rtsp av stream inside the center while if a bird is recognised the name of that bird appears on a screen (in Dutch)? A bit like the Merlin bird app sound ID changing colour when the sound re-appears. Also not storing the sounds because the sd-card will get full fast in our location.

  • @dannythed
    @dannythed Рік тому

    I'd love to get that chicken printed, is that a separate base attached?

  • @tomarmstrong8057
    @tomarmstrong8057 Рік тому +1

    Excellent clip ! Thanks for sharing. Just wondering if there is a way to use birdnet PI out in the field (away from wifi) and still maintain the ability to recieve sms alerts ? I would like to monitor the presence of a single species. Can 4g functionality be added and used for this purpose ? Any help appreciate. Cheers

    • @Core-Electronics
      @Core-Electronics  Рік тому

      I reckon you absolutely can! All the scripts are open source and ready to adapted to whatever you desire. There are some great USB 4G dongles that you could incorporate into the system to get internet connectivity. I have even done a guide on a 4G Hat for Raspberry Pi diving into Text message alerts. - ua-cam.com/video/ABnwz-IYzqA/v-deo.html.
      This system already makes it easy to isolate for one particular bird so you are almost there 😊 Kind regards - Tim

  • @johnmccormick2883
    @johnmccormick2883 2 роки тому +1

    Very interesting! Thanks for sharing!

  • @ThistleBeaksBirdCams
    @ThistleBeaksBirdCams 8 місяців тому

    Is it easy enough to use my rstp live stream mic with this? If so how? Thanks

  • @eugenn2848
    @eugenn2848 2 роки тому +2

    Need advice, please. Is it possible to build a sound recognition system based on raspberry pie that could be able to detect a dog's barking and perform/trigger some action afterwards? My neighbour's dog's barking drives me crazy, I would like to be able to automatically detect that barking and play some loud music in respond every time that animal is barking. Is it possible/difficult to accomplish?

    • @Core-Electronics
      @Core-Electronics  2 роки тому +1

      Absolutely Possible! And what a useful project it would be. Come to our forum and we'll be able to figure out the best hardware/way to tackle your idea 😊 - forum.core-electronics.com.au/

    • @eugenn2848
      @eugenn2848 2 роки тому +1

      @@Core-Electronics Thanks!

  • @NikoChefZocker
    @NikoChefZocker 2 роки тому +2

    you should mention that the website can be reached under a different URL if you change the hostname in the advanced settings when creating the image!!

  • @szabicsabai9765
    @szabicsabai9765 Рік тому

    Do u have any idea how to make it for musical instruments?

  • @growtocycle6992
    @growtocycle6992 6 місяців тому

    You said it's got a 75% confidence on a Maggpie... It recognized a kookaburra. Any human can recognize these very distinctive calls.
    Can you demonstrate it's use for more nuanced/less distinctive birdcalls?
    Thanks, this is s brilliant thing to make open source!

  • @theodorpintilie7008
    @theodorpintilie7008 2 роки тому +1

    Hello! Really nice video. Do you know how the data can be sent in a Grafana dashboard? I can't really figure out from the html code.

    • @Core-Electronics
      @Core-Electronics  2 роки тому

      Thank you kindly mate 😊 and sounds like an excellent question for the Github Discussion page - github.com/mcguirepr89/BirdNET-Pi/discussions

  • @dibyojyotibhattacherjee4279
    @dibyojyotibhattacherjee4279 2 роки тому +1

    some arduino and esp32 series pls, btw, awesome video...

  • @eszterkun857
    @eszterkun857 2 роки тому +1

    Great channel!

  • @ricki-bobby
    @ricki-bobby 2 роки тому +3

    Great use of a Pi. I'd like to take this concept to the next level and create a "set it and forget it" type setup in a waterproof case. Obviously that might take engineering a power source into this. The waterproof case it not an issue because there are lots of options for that. I'd want to be able to record only during certain hours and store everything on a large micro-SD. Has anyone gone down this path yet?

    • @Core-Electronics
      @Core-Electronics  2 роки тому +2

      Lots of BirdNet-Pi Projects up and running now come check some of their details out here 😊 github.com/mcguirepr89/BirdNET-Pi/discussions/69
      I know for sure that the best bird calls get stored on the Micro-SD by default, you may find if you store every noise you may end up filling your SD card quite quickly.

    • @Bluesbrother1988
      @Bluesbrother1988 Рік тому

      Did you ever continue with this idea? Had the same idea to drop it away from noise (e.g. deep into a forest).

    • @ricki-bobby
      @ricki-bobby Рік тому +1

      @@Bluesbrother1988 People steered me to the AudioMoth project. It seems to do exactly this as an all in on package. Unfortunately, it is unavailable for purchase because of a large backlog

  • @JimLyons
    @JimLyons 2 роки тому +1

    I just got mine going here in Boise, ID USA. I have already learned so much, as a long-time user of the Merlin and BirdNET mobile apps. Wondering about so many advanced elements of the system but getting some great results already. A simple question though - where do you download the recording for inclusion in an eBird list?

    • @Core-Electronics
      @Core-Electronics  2 роки тому

      Absolutely awesome getting your system running 😊 everything to know in regards to the TFLite BirdNet Analyzer (the backbone of this system) can be found here - github.com/kahst/BirdNET-Analyzer
      Not I'm not 100% sure what your question is. If you want to download the best recordings you can do so using the BirdNet-Pi UI, it'll be a little ellipsis symbol that you can click on. If you meant you wanted to ad new birds call to the system that would be difficult.

  • @coldhardwick
    @coldhardwick 2 роки тому +1

    Where'd you buy that pi?

  • @osama_alsafi
    @osama_alsafi 2 роки тому +1

    Another interesting project from Core electronics.
    Just wondering could we add an Audible output that calls out the name of the identified birds?

    • @Core-Electronics
      @Core-Electronics  2 роки тому

      Cheers for checking it out 😊 and you definitely could do that. Espeak is a TTS (Text To Speech) software which would work perfectly for the job. You would need to add some custom Python code and a speaker but with those two sorted your Raspberry Pi would be audibly identifying species.

  • @petersutcliffe4970
    @petersutcliffe4970 2 роки тому

    This project is just what I was looking for. Everything seemed to install ok. When I ran the web interface it appeared but nothing was being logged.
    When I looked at the log there were a number of errors. Starting the service gave an exec format error. Then main process exited with status 203/EXEC. There was also no such file or directory for /home/pi/BirdSongs/*/*Analyzed. I’m a newbie so any help would be appreciated.
    Thanks for posting this project.
    Peter

    • @petersutcliffe4970
      @petersutcliffe4970 2 роки тому +1

      Ok. Went to GitHub and used the links to uninstall and then reinstall. The errors have gone.

  • @maripro9555
    @maripro9555 Рік тому

    I found this project so exciting and got the hardware for it. However, I, as well as others recently get an error during installation. Are there people here with the same problem and/or a solution? The error is already addressed in the github project.

  • @kosmonautofficial296
    @kosmonautofficial296 2 роки тому +1

    this is so sick

  • @DJvvAZZ
    @DJvvAZZ 2 роки тому +1

    This is Soo cool!

  • @sophiahuang7025
    @sophiahuang7025 2 роки тому +1

    Nice videos

  • @christopherlyons7613
    @christopherlyons7613 2 роки тому +5

    Interesting but you left a lot of unanswered questions. More details would be good. Overall, how well does it work? What factor does the microphone used play into how well it works? What other microphones could lead to better results? How well can it recognize bird calla against normal background noises? Generally you will have several birds calling simultaneously, how well can it distinguish one call from others in the same sound sample? Did you actually try this in your own backyard? How well did it work? What did you see as the positives & negatives of this application? Thanks.

    • @DanielSMatthews
      @DanielSMatthews 2 роки тому +3

      There is a lot more information on _birdnet cornell edu_

    • @3nertia
      @3nertia 2 роки тому

      @@DanielSMatthews Doesn't seem like their site really details *any* of that though?

    • @DanielSMatthews
      @DanielSMatthews 2 роки тому

      @@3nertia You know how scientists put a list of publications at the end of their work, you know all that stuff you _never_ read, well perhaps you should...

    • @3nertia
      @3nertia 2 роки тому +1

      @@DanielSMatthews You're showing your ignorance - most scientific papers worth reading are behind paywalls lmfao
      Thanks for the laugh though, Dunning-Kruger!

    • @DanielSMatthews
      @DanielSMatthews 2 роки тому

      @@3nertia Then learn to jump higher, little troll.

  • @raytry69
    @raytry69 Рік тому

    6:46 Italian genes spotted!

  • @Redbeef
    @Redbeef 2 роки тому +1

    how is the thumbnail not Tim flapping 6:20

    • @_vicary
      @_vicary 2 роки тому +1

      I wanted to roast him for being cringe, but this comment stopped me. Oh wait...