Bounding Box Test and Training endpoint Exception about tiefvision

[ivancruzbht]

I just found out the issue, I have to manually create all the dest directories that this endpoint uses, i.e.

tiefvision/resources/bounding-boxes/crops
tiefvision/resources/bounding-boxes/crops/extended
tiefvision/resources/bounding-boxes/scaled-images

For a future feature, it would be nice to have the backend handle all of this, or maybe mention it in the md file :)

[thebearer696]

Hi @ivancruzbht I am facing an issue that might be related to yours, I created the directories and the endpoint works perfectly (at-least it says) but then upon running the file: encode-training-and-test-images.lua I get the following error log

~/tiefvision/src/torch/models/encoder.model Encoder loaded. nil ~/tiefvision/resources/bounding-boxes/0-train.txt: 0 nil ~/tiefvision/resources/bounding-boxes/1-train.txt: 0 Class 0 with files 0 luajit: bad argument #2 to '?' (out of bounds at ~/torch/pkg/torch/lib/TH/generic/THStorage.c:202) stack traceback: [C]: at 0x7f8ad2c9eb70 [C]: in function '__index' encode-training-and-test-images.lua:66: in function 'encodeData' encode-training-and-test-images.lua:80: in main chunk [C]: at 0x557a42ac5470
What I understand from this is that it is not able to locate 0-train.txt and 1-train.txt but I am unable to generate those. Any sort of help is appreciated.

[neerajvashistha]

Exactly same thing here 0-train.txt and 1-train.txt are not generated!!!, following the instruction given in developers guide, any help would be appreciated.

Thanks