OpenGVLab / OpenGVLab/InternVideo

Zero-shot retrieval reproduction issue

Open
#112 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
2.4k
Forks
160
PR merge metrics
No merged PRs in 30d

Description

According to the ReadMe at https://github.com/OpenGVLab/InternVideo/tree/main/InternVideo1/Downstream/Video-Text-Retrieval, the zero-shot retrieval results will be obtained after running the command ./zeroshot_scripts/eval_msrvtt.sh. This command will execute the main_task_retrieval.py. But in "main_task_retrieval.py", I find that the model is CLIP4CLIP, instead of ViCLIP. I'd like to know how to conduct zero-shot video-text retrieval experiments with pretrained ViCLIP.

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the README section at InternVideo1/Downstream/Video-Text-Retrieval and run or inspect zeroshot_scripts/eval_msrvtt.sh. Trace its call into main_task_retrieval.py, compare the referenced CLIP4CLIP path with the available ViCLIP materials, and document the verified procedure for running zero-shot video-text retrieval with pretrained ViCLIP.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning, search
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.