Intel® Distribution of OpenVINO™ Toolkit
Community support and discussions about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all things computer vision-related on Intel® platforms.
6064 Discussions

Error while running inference on a custom segmentation network

idata
Employee
482 Views

This is a simple semantic segmentation network that I'm trying to run on the movidius neural compute stick.

 

Compilation of graph was successful.

 

Input shape : (1, 240, 320, 3)

 

Output shape : (1, 240, 320, 2)

 

But when running the inference I got the following error.

 

E: [ 0] dispatcherEventReceive:200 dispatcherEventReceive() Read failed -1

 

E: [ 0] releasePacketFromStream:362 There is no packet to release

 

E: [ 0] dispatcherEventSend:730 Write failed -4

 

Could someone help me out with this issue? Thanks

0 Kudos
6 Replies
idata
Employee
221 Views

@abhi Can you provide a link to your model for testing? I've run into some issues with custom networks hanging the NCS device, but I would like to test this case to see if it is related. Thanks.

idata
Employee
221 Views
idata
Employee
221 Views

@abhi I want to let you know that I was able to reproduce the issue you reported here and we are taking a deeper look into the issue. Right now it does look like it is related to the custom Tiny Yolo v2 issue that some users have reported. I'll keep you updated. Thanks for your patience.

idata
Employee
221 Views

@Tome_at_Intel Thank you for the reply.

idata
Employee
221 Views

@Tome_at_Intel Any help regarding the issue would be highly appreciated. Thanks.

idata
Employee
221 Views

@abhi We're working on a release candidate right now to address the device disconnect issue with some custom networks, but at the moment we don't have a time frame for the release yet. We appreciate your patience while we work on this issue. Thank you very much.

Reply