search

ucbdrive / few-shot-object-detection

Implementations of few-shot object detection benchmarks
Apache License 2.0
1.08k stars 225 forks source link

Confusion about the split of test set(minival 5000) #110

Open tymanman opened 3 years ago

tymanman commented 3 years ago

I found that you give a split of minival set which is not consistent with the original split of official set.Why do u do like this,thank u .

xinw1012 commented 3 years ago

To clarify, which official set are you referring to? Neither COCO nor Pascal VOC is designed for few-shot object detection so we somehow need a manual split of the dataset.

Please clarify if the response is not what you are looking for. I'm not sure if I fully understand your question. Thanks!

tymanman commented 3 years ago

I mean there is already a minival split in COCO2017,which is not consistent with your manual split,actually.So I get confused about why to give a new manual split.In.other words,for some other consideration?

发自我的iPhone

------------------ Original ------------------ From: Xin Wang @.> Date: Fri,May 14,2021 10:38 AM To: ucbdrive/few-shot-object-detection @.> Cc: tymanman @.>, Author @.> Subject: Re: [ucbdrive/few-shot-object-detection] Confusion about the split of test set(minival 5000) (#110)

To clarify, which official set are you referring to? Neither COCO nor Pascal VOC is designed for few-shot object detection so we somehow need a manual split of the dataset.

Please clarify if the response is not what you are looking for. I'm not sure if I fully understand your question. Thanks!

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

Caltech-Z commented 1 year ago

I would like to ask, I have a dataset of my own, how should I proceed to divide it step by step into a dataset suitable for small sample target detection? Thank you very much, is there an introductory document on this?