Skip to main content

Frequently Asked Questions

This page covers common questions about downloading, starting, and connecting devices to Xense Studio. Select a question to expand its answer.

Download and Startup​

How do I get Xense Studio?

You can download Xense Studio from the official Xense website.

Why is the first launch slow?

On its first launch, the application initializes components, loads the inference model, and writes its cache. Subsequent launches are usually noticeably faster after initialization is complete.

Windows Environment​

What should I do if the GUI cannot start on Windows and displays an “access denied” or similar error?

Check whether the Windows Smart App Control feature is enabled. If it is enabled, open Windows Security, turn off Smart App Control, and then try starting the GUI again.

Smart App Control set to Off in Windows Security

Linux Environment​

How do I resolve a Linux Qt plugin error?

If you see Could not load the Qt platform plugin "xcb", run the following commands to install the missing dependency:

sudo apt-get update
sudo apt-get install libxcb-cursor0

Device Connection​

What should I do if the sensor list is empty?

A: Check system camera access permissions, USB permissions, sensor power, the configuration file directory, and the "Rescan" action. For remote devices, also check network connectivity and the compute unit service status.

What should I do if Xense Studio reports that the config file could not be read?

Contact the support team for a config file matching the sensor, then copy it to the localconfig folder in the Xense Studio installation directory. Xense Studio will prioritize this local configuration file.

The localconfig folder in the Xense Studio installation directory

Runtime Errors​

Why does the application crash after I connect a sensor?

If the application starts normally without a sensor but reports Segmentation Fault or stops near Init infer engine after a sensor is connected, check the NVIDIA driver first:

nvidia-smi

Then verify that CUDA, cuDNN, the SDK, and the GPU driver are compatible. You can temporarily switch to CPU inference to distinguish a GPU environment problem from a device communication problem.