Evaluate similarity or clustering algorithms for video files

[dennyabrain]

Evaluate similarity or clustering algorithms for video frames. Finalize Sampling Strategy for videos.

[Snehil-Shah]

Hi.

[aatmanvaidya]

• 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.

[aatmanvaidya]

@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

[aatmanvaidya]

• run video models on dataset and visualize using t-SNE, also do k-means clustering
• implement sampling strategies for video.

[aatmanvaidya]

• read and review sampling strategies
• Explore - closer frames as input, input higher moving frames etc.
• how to select better n key frames.

[aatmanvaidya]

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.