I'm trying to run the example python script that has come along with the repo. The input image as you can see in the source code is electric guitar. But following is the output that I'm getting after running the script.
Start download to NCS…
inception-v3 on NCS
(112, 'nematode, nematode worm, roundworm', 0.02176)
(550, 'envelope', 0.01718)
(624, 'letter opener, paper knife, paperknife', 0.012436)
(474, 'can opener, tin opener', 0.011734)
(513, 'corkscrew, bottle screw', 0.010895)
Can anyone please explain what could be the issue? Are there any pre-requisites before running the script?
I have not made any changes to this example. Has anyone else noticed the same behavior (Guitar being recognized with 2% probability) on NCS? Any thoughts on why this would be happening? Thanks !!
Found a solution to my problem.. The issue was, when i run the python run.py, I was getting incorrect results. But if I run the same code through 'make run' it is actually working fine with good prediction probability.