Hi everyone, after research on the latest DLStreamer repo on Github, I found a folder about license plate recognition using LPRNet: https://github.com/openvinotoolkit/open_model_zoo/tree/master//models/public/license-plate-recognition-barrier-0007
However, I can't find anything about the training task, maybe it only exists in the previous version (I also find on training_extension repo and have no result). My dataset only have numbers from 0 to 9 and Alphabet characters. How can I able to train that model?
Hi Hungtrieu07,
The license-plate-recognition-barrier-0007 model in the Open Model Zoo is provided only for inference, and unfortunately, the original training code is not available, even in the older training_extensions repo.
If you want to train a similar model using your own dataset (with digits 0–9 and alphabet characters), you can use an open-source PyTorch implementation of LPRNet, such as this one: https://github.com/sirius-ai/LPRNet_Pytorch (This is an external link and is not maintained by Intel)
Regards,
Aznie
連結已複製
Hi Hungtrieu07,
The license-plate-recognition-barrier-0007 model in the Open Model Zoo is provided only for inference, and unfortunately, the original training code is not available, even in the older training_extensions repo.
If you want to train a similar model using your own dataset (with digits 0–9 and alphabet characters), you can use an open-source PyTorch implementation of LPRNet, such as this one: https://github.com/sirius-ai/LPRNet_Pytorch (This is an external link and is not maintained by Intel)
Regards,
Aznie
