Dataset Size

[ly-muc]

Dear authors,

Thank you for your work. I have a question about your implementation.

In your paper you describe in the section "Datasets" that Cityscapes has 2975 sequences with 30 frames each, VIPER 134k frames and SYNTHIA-Seq 8000 frames.

However, your implementation indicates that a much smaller number of unique frames were used for training, i.e. 13k for VIPER and 849 frames for Synthia. Could you please elaborate on the reasoning behind this decision? Particularly as the difference between VIPER SYNTHIA is also significant.

Thanks for clarifying!