Evaluate similarity or clustering algorithms for video files
dennyabrain opened this issue · comments
Denny George commented
Evaluate similarity or clustering algorithms for video frames. Finalize Sampling Strategy for videos.
Snehil Shah commented
Hi.
Aatman Vaidya commented
- Try out more Video Models for Embedding
- Try self-supervised clustering algorithms
- Do a review of lit for video pre-processing. Read up on better sampling strategies.
Aatman Vaidya commented
@Snehil-Shah this is that kaggle competition I was talking about - https://www.kaggle.com/c/nfl-health-and-safety-helmet-assignment/discussion/285112
they have used k-means clustering for a different use case, do see if this helps you out
Aatman Vaidya commented
- run video models on dataset and visualize using t-SNE, also do k-means clustering
- implement sampling strategies for video.
Aatman Vaidya commented
- read and review sampling strategies
- Explore - closer frames as input, input higher moving frames etc.
- how to select better n key frames.
Aatman Vaidya commented
Notes - Jul 15
- extract keyframes from ffmpeg and then do QR
TODO
- Profile the CLIP embedding model.
- keep experimenting with more sampling strategies.
- collect dataset of low-quality videos in Indian Languages.