tiny yolo segmentation fault on tk1

Autonomous Machines Jetson & Embedded Systems Jetson TK1
1

i have compiled the yolo using the following options:
GPU=1 &CUDNN=0 &OPENCV=1

but unfortunately while i am running it ,it shows me the detected objects then a segmentation fault!

Any help?

2

Check dmesg before and after the seg fault. Post any error output (use the code block icon in the upper right with “</>” around the dmesg quote).

3

Hi,

Thanks for your question.
Could you share more error log of your issue. (Please check dmesg as linuxdev said)

By the way, we have enable YOLO with tensorRT. Read more in this topic:
https://devtalk.nvidia.com/default/topic/990426/jetson-tx1/tensorrt-yolo-inference-error/

4

Hi eslam_bakr,

Have you clarified and resolved this problem?
Any update for this issue?

Thanks

5

yes thanks a lot ,i solved it by a very strange way
when i make the debug flag in the make file=1 to debug the error it works !

by the way i want to know if i want to buy 100 thousand of tegra tk1 (not thw whole kite)
how much it will be?

thanks in advance

6

Hi eslam_bakr,

Please contact with your region sales for your requirement or send inquiry through below window:
[url]https://developer.nvidia.com/contact[/url]

Thanks

7

unfortunately i am in Egypt and there is no sales for NVIDIA here,if there is and i dont know please tell me ?

and i send inquiry through your link but there is no answer till now

thanks

8

Hi eslam_bakr,

Already highlighted to relevant team to follow up this sales inquiry, you should be able to receive the response soon.

Thanks