[optimalsolver]

Rather than doing self-supervised learning on the actual video frames, why not do it on the byte sequence that represents the video file?

[mkaic]

You might find this paper interesting: [JPEG-LM: LLMs as Image Generators with Canonical Codec Representations](https://arxiv.org/abs/2408.08459)

[optimalsolver]

Thanks. This is exactly the kind of thing I was looking for.