Hi
Actually, I have a problem about image size. When I used a big image, it was fail on EVM, but it was OK on HOST.
I tried with a small image, and then It was OK on EVM. It looks like there is relation in image size.
So, I would like to ask a caffe model of TI samples.
As you know, Ti used tidl_net_onnx_tiadsegNet_v2 for semantic segmentation. But I couldn't get a caffe model and prototxt of that network.
If I have those files, I think I can get more information through importing.
I was wondering If I could get a caffe model and prototxt of 'MobileNet-1.0V2 + ASSP'.
Best regards
Yongsig.