AI Listener

AI Listener consists of two types. One is the general Command process listener; the other is the Lobe tool template.

Command Listener allow user to broadcast the zip file to the remote screens. And the CMPlayer will monitor and listen to the standard output of the command.

Users should provide the Zip file, and the command text to run under the zip folder i.e. Python Main.py. The command can be AI model runner or any other applications which can generate standard output.

User can set up the possible outputs from the command. Every line in the command outputs is regarded as a new output. They will be used for conditional playback and appear in the dropdown list to match the condition.

To download the Lobe AI app, use the following link:

https://www.ryarc.net/Release/Lobe.exe

AI Templates use AI models with its creation, training and deployment of machine learning algorithms that emulate logical decision-making based on available data. The AI models that we use are created and trained using the Mirosoft Lobe application. https://www.lobe.ai/ Once the model is 100% trained, the .json and .onnx files can be exported.

Login to CMService as Domain Admin and from left navigation menu open ‘AI Listeners’.

Create New AI Listener.

In AI Listener you can import the .onnx file in the Module File section, and the .json file in the Description File section. (from the previously trained module from Lobe).

There are two ways to use the camera for AI Listeners.

Open Linux CMPlayer Configuration file and go to General Settings.

1. Use Web Camera directly.

2. Use Rtsp Camera. Run the Rtsp.exe and the URL from the Command Prompt enter in the ‘Capture the URL instead of the local camera’ field, save the changes and Run Diagnostics.

There are two ways to use camera for AI Listeners.

Open Windows CMPlayer Configuration file and go to General Settings.

1. Use Web Camera directly.

2. Use Rtsp Camera. Run the Rtsp.exe and the URL from the Command Prompt enter in the ‘Capture the URL instead of the local camera’ field, save the changes and Run Diagnostics.

Note: The Lobe AI Templates for Windows work only on CMPlayer x64, while Command Listener works for both x64 and x86.

Login to CMService as Domain Admin and open Campaigns. Import some media to the campaign.

Right click on media you want to add the condition Playlist Settings > Conditional.

In the modal window click on AI Match and ‘Add new AI condition’ button. New modal window with list of all created AI Listener will be shown. Choose one of the models and click on ‘OK’ button.

The AI model will be added with two dropdown fields, first one, for choosing operators:

Equal (= ), Less than (<), Greater than (>) and Not equal (<>), and the second one for choosing model.

Choose the condition and click on ‘OK’ button.

In each media that contains condition play, there will be an information icon about the condition.

Note: Each media can have more than one AI condition and the condition can be added on different types of media.

Create one Playlist and add different types of media. Right click on media you want to add the condition and click on the ‘Conditional’.

The Conditional Play modal window will be shown. Choose the AI Match condition and click on OK button. There will be an information icon near the media name about the condition.

The AI condition can be used on entities level as Campaigns, Tickers and Commands.

Open one of the entities, go to Settings tab and tick the ‘Play only when the condition matches’ option. The modal window for Conditional Play will be opened.

Once the player starts playing, the camera will turn on. The media/campaign will play only when the conditions match.

Command listener will make sure the zip command keeps running as described in the CMService Web UI. It recovers the command process if no output received within 10 minutes.

The command or AI output is recorded in the player log files as follows:

Log Entry : 4:44:26 PM Tuesday, 30 September 2025

:
:Command AI: 1 Blue LED

——————————-

Log Entry : 4:44:27 PM Tuesday, 30 September 2025

:
:Command AI: 3 Blue LEDs