some questions about training input

liuruijin17 / LSTR

This is an official repository of End-to-end Lane Shape Prediction with Transformers.

BSD 3-Clause "New" or "Revised" License

644 stars 130 forks source link

some questions about training input #90

Open kidcats opened 2 years ago

kidcats commented 2 years ago

Thank you for your very good work! now i hava one question.

    def forward(self, iteration, save, viz_split,
                xs, ys, **kwargs):

i know xs is original image and mask,but i can not understand ys mean what. and i see ys in calculate loss as gt.
could you tell more detail about dateset pre-processing?
thanks