Forums - yolo post process and yolo to dlc conversion

3 posts / 0 new
or Register
Last post
yolo post process and yolo to dlc conversion
ashokelse
Join Date: 20 May 22
Posts: 1
Posted: Fri, 2022-05-20 03:54
Top

I converted tensorflow  pb to dlc model And i'm using qualcomm Neural processing SDK.

I'm trying to run object detection model

Problems:

  1. I have three output layers (yolo ) but i'm able to load only one ouput layer, In dlc model i checked the info., it has three output layers
  2. And i'm getting this error:


E/AndroidRuntime: FATAL EXCEPTION: AsyncTask #1
    Process: com.qualcomm.qti.snpe.imageclassifiers, PID: 5385
    java.lang.RuntimeException: An error occurred while executing doInBackground()
        at android.os.AsyncTask$4.done(AsyncTask.java:399)
        at java.util.concurrent.FutureTask.finishCompletion(FutureTask.java:383)
        at java.util.concurrent.FutureTask.setException(FutureTask.java:252)
        at java.util.concurrent.FutureTask.run(FutureTask.java:271)
        at android.os.AsyncTask$SerialExecutor$1.run(AsyncTask.java:289)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1167)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:641)
        at java.lang.Thread.run(Thread.java:919)
     Caused by: java.lang.ArrayIndexOutOfBoundsException: length=10; index=161101
        at com.qualcomm.qti.snpe.imageclassifiers.tasks.ClassifyImageWithFloatTensorTask.doInBackground(ClassifyImageWithFloatTensorTask.java:70)
        at com.qualcomm.qti.snpe.imageclassifiers.tasks.ClassifyImageWithFloatTensorTask.doInBackground(ClassifyImageWithFloatTensorTask.java:24)
        at android.os.AsyncTask$3.call(AsyncTask.java:378)
        at java.util.concurrent.FutureTask.run(FutureTask.java:266)
        at android.os.AsyncTask$SerialExecutor$1.run(AsyncTask.java:289) 
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1167) 
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:641) 
        at java.lang.Thread.run(Thread.java:919) 
I/Process: Sending signal. PID: 5385 SIG: 9
3.  i have problem in ClassifyImageWithFloatTensorTask.java  this file and doInBackground
 


for (Pair<Integer, Float> pair : topK(1, array)) {
    result.add(mModel.labels[pair.first]);
    result.add(String.valueOf(pair.second));
}
 
  • Up0
  • Down0
 
 Re: yolo post process and yolo to dlc conversion 
 weihuan
 Join Date: 12 Apr 20
 Posts: 270
 Posted: Sat, 2022-05-21 20:13
 Top
 
Dear customer,
Could you please try to run this issue model with snpe-net-run to get the results you want?
BR.
Wei
  • Up0
  • Down0
 
 Re: yolo post process and yolo to dlc conversion 
 balavignesh.g
 Join Date: 12 Apr 22
 Posts: 3
 Posted: Tue, 2022-05-24 04:43
 Top
 
Could you run Yolo in Qualcomm? If so, what are files need to run using Gstreamer?
 
  • 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.