Nvinfer on metadata

Please provide complete information as applicable to your setup.

**• Hardware Platform: **Jetson
**• DeepStream Version: ** 6.3
• JetPack Version: 5.1.2
**• TensorRT Version: ** 8.5.2
**• NVIDIA GPU Driver Version: ** R35
• Issue Type: questions or new requirements

I have pipeline like this:
rtspsrc → rtph264depay → nvv4l2decoder-> nvstreammux → primary-gie → dsexample → secondary-gie1 → detectcolors (based on dsexample) …
I have a list of floats at the end of detectcolors that I can add to metadata.
Would it be possible for me to run my tensortrt model with nvinfer on this list of floats (not on the image data, only on these values of the metadata)?

Thanks

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

Yes. It is possible. You need to prepare the floats list in tensor meta NvDsPreProcessBatchMeta with nvpreprocess (Gst-nvdspreprocess (Alpha) — DeepStream documentation 6.4 documentation), then you can use the “input-tensor-from-meta” property to get the tensor meta into nvinfer.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.