OpenGVLab / InternVideo

[ECCV2024] Video Foundation Models & Data for Multimodal Understanding
Apache License 2.0
1.31k stars 85 forks source link

LSMDC Annotation Files #137

Closed fmthoker closed 3 months ago

fmthoker commented 3 months ago

Dear authors, can you share the annotation files that are used for LSMDC retrieval? I am trying to reproduce the results using ViClip models but it's not clear which annotation files were used. Thanks in advance.

JingXiaolun commented 3 months ago

@fmthoker ,sorry to disturb you. When I reproduce the LSMDC results, I find the MeanR is always around 200, I use the 2016 splits (101046 for training, 7408 for validation , and 1000 for test), and the video is compressed to 3fps (224x224). Do you encounter similar problems? Looking forward to your reply.

fmthoker commented 3 months ago

@JingXiaolun no i don't have such issues. Please check this issue for my lsmdc performance