Forums - Loading non-image inputs to snpe-net-run to run model

3 posts / 0 new
Last post
Loading non-image inputs to snpe-net-run to run model
22imonreal
Join Date: 10 Feb 21
Posts: 80
Posted: Wed, 2021-05-26 05:15

Hello,

I have followed the snpe-net-run examples and my understanding is that snpe-net-run is input agnostic, meaning we could define any input we wanted and not just gbr raw files as per the esamples. I have tried successfully with a 1D vector, but would now like to load a 640,320,6 channel deep tensor as an input as well as two other 640,320,1 grey scale images. How would I be able to express these inputs (originally created as Python ndarrays) and write them to files that snpe-net-run could interpret and run the model on? For example, I cannot write the 6 channel deep input as a GBR raw image file, and neither a a txt file which is only 2D, so how would one go about it?

Thanks

  • Up0
  • Down0
kevin.dai
Join Date: 21 Oct 20
Posts: 137
Posted: Wed, 2021-05-26 10:35

Let us talk Snpe expert and give you update.

Thanks

  • Up0
  • Down0
kevin.dai
Join Date: 21 Oct 20
Posts: 137
Posted: Wed, 2021-07-14 15:17

Hi 22imonrea

Yes, you can applied SNPE for non-image inputs, but your model should support your input, For example, if your input is voice, your model should key word detecytion model. make sure model to match your input type.

Thanks

Kevin

 

  • Up0
  • Down0
or Register

Opinions expressed in the content posted here are the personal opinions of the original authors, and do not necessarily reflect those of Qualcomm Incorporated or its subsidiaries (“Qualcomm”). The content is provided for informational purposes only and is not meant to be an endorsement or representation by Qualcomm or any other party. This site may also provide links or references to non-Qualcomm sites and resources. Qualcomm makes no representations, warranties, or other commitments whatsoever about any non-Qualcomm sites or third-party resources that may be referenced, accessible from, or linked to this site.