Thanks very much, Trevor

That sounds really promising.
As you say, there are good databases of birds -- xeno-canto is actually pretty global in scope.
Other sounds are probably less well organised.

We have contacts on this topic. It might be interesting and useful to think how they might be in touch - some kind of seminar in real-time ecoacoustic  analysis by AI?

bw
g


On 24 Aug 2021, at 16:14, Trevor Hebert <thebert@stanford.edu> wrote:

I'm currently working with some CS students on a somewhat similar system, except it monitors and detects wildfires. That effort should produce a pretty good conceptual framework and probably some useful code as well. The hardest part will be to acquire sets of training data to teach the AI how to identify sounds. Some areas, like North America and Europe should already have pretty good libraries of bird sounds. Insects? Amphibians?  Mammals? Anthropogenic sounds? Maybe less so. 

Integration with Locus Sonus would be really cool - I could envision a scrolling sidebar on the world map that displays currently detected species that link the the audio streamer location. A searchable archive would be useful to researchers.

Trevor

Trevor Hebert
Jasper Ridge Biological Preserve
Stanford University
650-851-6815


From: Grant Smith <grant@soundtent.org>
Sent: Monday, August 23, 2021 10:36 AM
To: Trevor Hebert <thebert@stanford.edu>
Cc: Peter Sinclair <petesinc2@gmail.com>; locustream@locusonus.org <locustream@locusonus.org>
Subject: Re: Open Microphones for IUCN World Conservation Congress
 
We re super interested to do this, and could potentially loop in some researchers from UK and elsewhere.

Some progress has been made on this already in various places, so we'd have some starting points, I think

Did you have thoughts on how to take it forward, Trevor!
Maybe a call would be useful

bw
g



On 23 Aug 2021, at 18:11, Trevor Hebert <thebert@stanford.edu> wrote:

On a different topic ... is there any interest in developing an AI engine for identification of sounds on our various streams? Perhaps integrated into Locus Sonus?

I probably can recruit some Stanford computer science students to work on this.

Trevor Hebert
Jasper Ridge Biological Preserve
Stanford University
650-851-6815


From: Grant Smith <grant@soundtent.org>
Sent: Monday, August 23, 2021 9:18 AM
To: Peter Sinclair <petesinc2@gmail.com>
Cc: Trevor Hebert <thebert@stanford.edu>; locustream@locusonus.org <locustream@locusonus.org>
Subject: Re: Open Microphones for IUCN World Conservation Congress
 
Thanks very much, Peter

The server is, of course, generally highly reliable and gives great service to ppl working with live environmental sound and related projects.

Of course we know now for the next Reveil event at start of May, so we can synch up to avoid issues. 

For one thing, from the Soundcamp side, we will review the design of the Reveil platform to reduce unnecessary demands on the server, which we didn't think through / anticipate.

bw
grant


On 20 Aug 2021, at 12:37, Peter Sinclair <petesinc2@gmail.com> wrote:

Hi Trevor and thanks for getting back to us. 

Concerning the issues during Reveil, thank you for pointing this out and yes we are aware that there where problems and apologise if it spoiled the event for anyone. It was due to two factors - one is that the service provider has limited the traffic to the locus mountpoint - we were not aware of this before (possibly becasue we had never hit the limit ) but it is rectifiable by announcing upcoming events to Creacast.com. The other problem is due to the architecture of the timeline page for Reveil which means that everytime someone updates the page it pings all the mountpoints including those which are not active because destined to be used for a performance at a specific time. We will fix these issues before the next soundcamp reveil. In the meantime it shouldn't be a problem for the soundwalk we are planning which is spread over several weeks (we are not expecting thousands of users at the same time).  

Thanks for you help,

Best
Peter

Le 19/08/2021 à 17:21, Trevor Hebert a écrit :
Hi Peter,
Our mic is always on and always happy participate in streaming events:

I've noticed during the last two Reveil events that LocusSonus has some kind of performance/capacity problem that impacts some streams, including ours. Our audio stream dropped out constantly during the Reveil event, something that never happens on our stream outside of Reveil. Is this a known issue? Too much traffic at once? It's not a network connectivity or equipment issue on our end.

thanks,
Trevor

Trevor Hebert
Jasper Ridge Biological Preserve
Stanford University
650-851-6815


From: Peter Sinclair <petesinc2@gmail.com>
Sent: Thursday, August 19, 2021 7:45 AM
To: kaorusaito <kaoru@cyberforest.jp>; teum11@gmail.com <teum11@gmail.com>;SarahD@cumbriawildlifetrust.org.uk <SarahD@cumbriawildlifetrust.org.uk>; grant@soundtent.org<grant@soundtent.org>; Trevor Hebert <thebert@stanford.edu>; Leah Barclay <leahbarclay@me.com>; Sukanta Majumdar <sukantamajumdar77@gmail.com>; mkin@centrum.cz <mkin@centrum.cz>; gangil yi<giy.hands@gmail.com>
Cc: Thierry BOTTI <thierry.botti@osupytheas.fr>; gregoire lauvin <greglauvin@gregoirelauvin.net>
Subject: Open Microphones for IUCN World Conservation Congress
 
Dear Streamers,

I hope you are all well and that you haven't been too badly effected by 
the pandemic.

The last presentation of the geolocated stream soundwalk "des Mondes au 
creux de l'oreille" that your stream participated in last year did take 
place, but was of course affected by lockdowns etc. The second event it 
was to be attached to - The World Conservation Conference 
https://www.iucncongress2020.org/ was postponed and is finally taking 
place this coming september.
We have decided to reactivate the piece for the event.

Would it be possible for you to switch on your microphone for the event 
(as soon as possible so that we can test the setup)? If this is not 
possible please let us know so that we can find and alternative.

As always our warm thanks for your participation!

Best Wishes
Peter