Support video retrieval dataset MSVD 支持 MSVD 视频检索数据集 #545
OpenMMLab-Assistant-004
started this conversation in
MMAction2
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
Support video retrieval dataset MSVD
支持 MSVD 视频检索数据集
Support MSVD video retrieval dataset in mmaction2, including the original dataset download and processing scripts (the processed data should be located in the data/video_retrieval/msvd directory and conform to the mmaction2 video retrieval task data format, which can refer to the already supported MST-VTT dataset), download and processing scripts can refer to prepare_msrvtt.py and prepare_msrvtt.sh in tools/data/video_retrieval, and need to update the README under tools/data/video_retrieval, mainly including the dataset citation and introduction, the invocation method of download and processing scripts, and the data directory structure tree after processing.
在 mmaction2 中支持 MSVD 视频检索数据集,包括原始数据集下载和处理脚本(经过处理后的数据应位于 data/video_retrieval/msvd 目录下且符合 mmaction2 视频检索任务数据格式,可参考已经支持的 MST-VTT 数据集),下载和处理脚本可参考 tools/data/video_retrieval 下的 prepare_msrvtt.py 和 prepare_msrvtt.sh,并且需要更新 tools/data/video_retrieval 下的 README,主要包括主要包括数据集的引用和简介、下载处理脚本的调用方法和处理完成的数据目录结构树。
Task Information
Repository:MMAction2
Difficulty:Medium
Points:20
Estimated Time Consumption:14 days
Task Status
Claimant's Github ID:xiaoyi0919
Claim Time:2023年07月26日
PR Link:open-mmlab/mmaction2#2622
Points to Note
❗️请注意:此任务属于超级视客营任务,不能被领取!请登陆 超级视客营平台 按照流程申请任务,我们会尽快审核~
Beta Was this translation helpful? Give feedback.
All reactions