PoseCNN
PoseCNN copied to clipboard
yuxng
how to download YCB-Video dataset
I have tried many ways to download YCB-Video dataset from Google drive, but it seems the file is too large to be download, I meet some network error and connection error. could anybody tell me how which tool will be better to download this large file? thanks for help！
@Abdul-Mukit can you upload several dataset, just a very small part of dataset, ten groups of images maybe ok? the network speed is very slow in my country. But I just want to see the training process, so don't need the whole dataset. could you help me? thanks a lot !!!
Ok, I'll do something and let you know. I actually may have corrupted the dataset last night. Was trying to move it from one drive to another but something went wrong. Anyway, I'll check today and let you know. Bye the way, have you tried DOPE? I am moving on to DOPE now. The documentation for creating new dataset seems doable which is very important for me.
@Abdul-Mukit thanks a lot !
do you mean Deep Object Pose Estimation (DOPE) – ROS inference (CoRL 2018)？ i don't try it. but thanks for information, but I will have a try
Hi, I uploaded some files of the dataset here. https://github.com/Abdul-Mukit/ycb_video_data_share One problem is that I couldn't upload the "models" folder as it was too big to upload. For the "data" folder, I uploaded one out of 92 folders: "0000". Inside it, I kept ten sets of data. I tried to keep some data in the data_syn folder as well. Other than, "data, data_syn and models" folder, all other folders should be just as they are in the original upload. Hope this helps.
Bye the way, I think I understand you download problem. I tried to download the dataset again yesterday myself. It stopped due to some network error and prompted a message that I can't download the dataset as too many people read/downloaded the file. It said that I should reset in 24 hours, but if not then I should contact my network administrator.
And yes, I was talking about Deep Object Pose Estimation (DOPE). It is better than PoseCNN although it is trained entirely on synthetic data.
I've uploaded YCB to Baidu NetDisk.
download link: https://pan.baidu.com/s/1FG7_wrNbBdFcJmh1UxuDFg passcode: 52xx
It has been split into small zip files, some folders are split into subfolders, and you need to merge them after unzip.
I've uploaded YCB to Baidu NetDisk.
download link: https://pan.baidu.com/s/1FG7_wrNbBdFcJmh1UxuDFg passcode: 52xx
It has been split into small zip files, some folders are split into subfolders, and you need to merge them after unzip.
Thanks for helping Chinese, maybe as well as North Korean, researchers! It'll be excellent if you can also put them on some IPv6 pt station such as byr neu6 :-)
@wangg12 Thank you for upload the dateset. I have a question about training. Did you put the whole dataset into the LOV folder or just create a symlink for "data" and "models" as the instruction in README.md ?
I've uploaded YCB to Baidu NetDisk.
download link: https://pan.baidu.com/s/1FG7_wrNbBdFcJmh1UxuDFg passcode: 52xx
It has been split into small zip files, some folders are split into subfolders, and you need to merge them after unzip.
Baidu is impossible to sign up for from the UK, as it does not accept UK phone numbers. And you can't download without an account. Could someone else mirror on a different site? Or maybe create a torrent?
I've uploaded YCB to Baidu NetDisk. download link: https://pan.baidu.com/s/1FG7_wrNbBdFcJmh1UxuDFg passcode: 52xx It has been split into small zip files, some folders are split into subfolders, and you need to merge them after unzip.
Baidu is impossible to sign up for from the UK, as it does not accept UK phone numbers. And you can't download without an account. Could someone else mirror on a different site? Or maybe create a torrent?
You can download Baidu NetDisk files on https://pan.baiduwp.com/, first blank is download link and second is passcode, but it seems that you must download these files one by one, not the whole folder together.
I've uploaded YCB to Baidu NetDisk.
download link: https://pan.baidu.com/s/1FG7_wrNbBdFcJmh1UxuDFg passcode: 52xx
It has been split into small zip files, some folders are split into subfolders, and you need to merge them after unzip.
Thanks for sharing, but Baidu downloader is too slow, could you share them on IPV6 resource website, like http://bt.neu6.edu.cn and http://pt.zhixing.bjtu.edu.cn, it'll be an awsome work!
I've uploaded YCB to Baidu NetDisk. download link: https://pan.baidu.com/s/1FG7_wrNbBdFcJmh1UxuDFg passcode: 52xx It has been split into small zip files, some folders are split into subfolders, and you need to merge them after unzip.
Thanks for sharing, but Baidu downloader is too slow, could you share them on IPV6 resource website, like http://bt.neu6.edu.cn and http://pt.zhixing.bjtu.edu.cn, it'll be an awsome work!
You could try pandownload without SVIP. It could reach about 2MB/s speed, or 10MB/s with SVIP.
@wangg12 OK, maybe you can try it if you are free. There are some CV dataset on PT web, such as ImageNet, COCO..., many Chinese university can use IPV6 download these resource, the speed is fast, more than 30MB/s and totally free. Anyway, thanks! :-)
I've uploaded YCB to Baidu NetDisk. download link: https://pan.baidu.com/s/1FG7_wrNbBdFcJmh1UxuDFg passcode: 52xx It has been split into small zip files, some folders are split into subfolders, and you need to merge them after unzip.
Thanks for sharing, but Baidu downloader is too slow, could you share them on IPV6 resource website, like http://bt.neu6.edu.cn and http://pt.zhixing.bjtu.edu.cn, it'll be an awsome work!
You could try pandownload without SVIP. It could reach about 2MB/s speed, or 10MB/s with SVIP.
Thanks! I'll try.
@wangg12 OK, maybe you can try it if you are free. There are some CV dataset on PT web, such as ImageNet, COCO..., many Chinese university can use IPV6 download these resource, the speed is fast, more than 30MB/s and totally free. Anyway, thanks! :-)
OK, I'd like try to upload dataset to some PT site. But my neu6 account level is not enough to upload anything. I think it could take me some time to figure out how to upload on ne6u.
@wangg12 OK, maybe you can try it if you are free. There are some CV dataset on PT web, such as ImageNet, COCO..., many Chinese university can use IPV6 download these resource, the speed is fast, more than 30MB/s and totally free. Anyway, thanks! :-)
I create a torrent but I don't know where to post it. The manager of ne6u didn't login since 5 days ago. And my torrent could not be seen by normal user until a moderator move it to public channel.
@YuiTH I found your PT on neu6, it's a nice job even it need some time to be approved. By the way, I have downloaded the YCB dataset using pandownload, it's truly a good downloader! Thanks!
I've uploaded YCB to Baidu NetDisk.
download link: https://pan.baidu.com/s/1FG7_wrNbBdFcJmh1UxuDFg passcode: 52xx
It has been split into small zip files, some folders are split into subfolders, and you need to merge them after unzip.
Thanks a lot for what u did! I've tried many times and several days for downloading it. : ( Thanks again! 👍
If anyone is looking to download the full YCB-V dataset we have a hosted version in S3 and can give you access. Just reach out to [email protected] with subject : YCB-V download
you can try this mirror https://nozomi.sk/ycb-video-dataset-download-mirror/
@mydansun Hi Daniel. I download "YCB_dataset.7z.030" file, can you please send more details how to extract data from that file.
@mydansun Hi Daniel. I download "YCB_dataset.7z.030" file, can you please send more details how to extract data from that file.
You need to download all splits and then use the command 7z x -tsplit YCB_dataset.7z.001 to uncompress.
@mydansun, I used the command 7z x -tsplit YCB_dataset.7z.001 (after downloading all the 30 volumes from one drive) and obtained a YCB_dataset.7z file. However, I am unable to extract contents from this 7z file. It says the End-of-central-directory signature not found. Could you please let me know how I can proceed further to use the dataset? I tried debugging this problem, but could not find a way to resolve this issue.
@Kowndinya2000 Could you check hash results to check if any files are damaged?
