can not download scannet? about fpconv

[lyqun]

Yes, only _vh_clean_2.ply and _vh_clean_2.labels.ply are used for data pre-processing.

Actually, we also provide a pre-processed ScanNet dataset used only for this repo, which you could download from Google Drive (please refer to README).

[longmalongma]

How big they are？

…

------------------ Original ------------------ From: Yiqun Lin <[email protected]&gt; Date: Wed,Sep 9,2020 10:52 AM To: lyqun/FPConv <[email protected]&gt; Cc: longma <[email protected]&gt;, Author <[email protected]&gt; Subject: Re: [lyqun/FPConv] can not download scannet? (#8) Yes, only _vh_clean_2.ply and _vh_clean_2.labels.ply are used for data pre-processing. Actually, we also provide a pre-processed ScanNet dataset used only for this repo, which you could download from Google Drive (please refer to README). — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

[lyqun]

9 GB totally for preprocessed pickles (3 files for train, eval and test splits).

About 20 GB (or maybe larger) for the official dataset ( _vh_clean_2.ply and _vh_clean_2.labels.ply only).

[longmalongma]

_vh_clean_2.ply and _vh_clean_2.labels.ply

Thanks for your reply!