Hi,I need ideas/help for a nice project:
It is a tour guide or simultaneous translation tool.
The idea is:
a) Design a simple webpage on a raspery or esp or something.
b) integrate a kind of stream player that will be connected to a live recording.
c) people within the same wifi network can connect to the local webpage (maybe with a little code…) and listen to the live translation of the speaker.
Of course… important low latency
::::::::::::
So the basic Idea is:
listener can join the translation via mobilephone WITHOUT installing an app!
The translation is NOT via Internet access, but via localhost webpage.
…
Any Idea of steps / apps with webpage to use in order to get such a project running?
one idea is to use obs and live stream to my pc…
but how to get visitors to have access wie WiFi in the room to my page?
It should not be Intenet…
so localhost or so…
here I am stuck a bit…
Well you should be able to do it like any live streaming platform, the only thing different is that only local network people can reach the page. With the web server on the machine, set it up to only accept requests from the local network range of IP addresses. Your guests on the network can then reach your computer through your IP address. For apache you can do things like suggested in this post…
With it, you can broadcast everything you need and you can set local access to only one IP, then use a VPN so that users of only a certain network have access. On the other hand, during installation, you specify the domain name or IP and add a password, and until the user enters the password, they will not be able to access your server
In any case, it is a convenient tool for your work. There are videos on YouTube that will show how to install it and access it, I would post the video here, but this is not possible, so I will just post commands to you.
Thank you for help.
I installed icecast2 and it works,But I have a 2 sec latence.
so i will do some tests if i can get less then 1 sec.
One solution seems to be sonobus…
havn,t tested that…
…
the project is for live simultaneous translation.
visitors should login the local WiFi.
Then with Qr code pointing to to localhost/translation…
here should be the live voice…
To create a local tour guide or simultaneous translation tool, follow these steps:
Hardware Setup:
Use a Raspberry Pi (or ESP32) with a microphone to capture audio and serve a webpage.
Optionally, connect speakers for local audio broadcasting.
Webpage Design:
Use a web server (Nginx/Apache) to serve a simple HTML5 page with an audio player that streams live audio.
Provide a QR code to help users connect to the local webpage via their mobile devices.
Live Audio Streaming:
Stream the live audio using Icecast or WebRTC for low-latency transmission.
Use Opus codec for efficient audio streaming.
Mobile Access:
Users connect to the local Wi-Fi network and open the webpage in their mobile browser to listen to the live stream without needing an app.
Low Latency:
Minimize buffering by configuring small buffer sizes in the stream and consider using WebRTC for optimal low-latency performance.
This setup will allow users to access the live translation or tour guide audio via a local network, ensuring real-time streaming with no app installation required.
Choose the Right Software: Use a software or platform that supports live voice translation, such as Google Translate, Microsoft Translator, or specialized translation software like Interprefy or KUDO for professional setups.
Set Up Audio Equipment: Ensure you have high-quality microphones, speakers, and a sound system for capturing and broadcasting the audio clearly. A headset with a microphone is ideal for individual translators.
Install the Translation Software: Install and configure the translation software on your computer or mobile device. Set the source and target languages for real-time translation.
Ensure Stable Internet Connection: A fast and reliable internet connection is crucial for seamless translation and minimal latency.
Prepare Your Translation Team: If using human translators, set up a system where each translator can listen to the source language and speak the translated version. You can use multiple channels for different languages if needed.