search

umautobots / vod-converter

Convert between visual object detection datasets
MIT License
307 stars 93 forks source link

Input for --from path and --to path #9

Closed benn94 closed 7 years ago

benn94 commented 7 years ago

Could you tell me what is the input for --from path and --to path? is it the whole dataset + annotation file? or is it just annotation file?

krosaen commented 7 years ago

these are intended to be paths to the root of the dataset

benn94 commented 7 years ago

@krosaen am i supposed to put only the path to annotation files?

krosaen commented 7 years ago

the root directory typically contains both annotations and images