jerryli27
commented
1 year ago I can't remember exactly since it's been a while. I believe it was on the order of O(1000).
It took me about 40-80 hours of work to create tools to label such a dataset and to do the labeling manually. the question of "whether it is worth it" depends on what you want to do exactly : )
You might be interested in this larger superset: https://github.com/SkyTNT/anime-segmentation. I haven't tried it out myself, so it's just an FYI.
On Fri, Dec 1, 2023 at 2:37 AM Arthur Wolf @.***> wrote:
Hello.
How many images were in the dataset you used to train? I want to know if its worth it for me to try to make a bigger one, or if it's just too big a task.
Thanks!
— Reply to this email directly, view it on GitHub https://github.com/jerryli27/AniSeg/issues/13, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACWYNYHC33Q7QDVYSI7JADTYHE7HPAVCNFSM6AAAAABACE6JDWVHI2DSMVQWIX3LMV43ASLTON2WKOZSGAYTSOJTG4ZDGNQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>
Hello.
How many images were in the dataset you used to train? I want to know if its worth it for me to try to make a bigger one, or if it's just too big a task.
Thanks!