hey HN, We're all NLP & CV engineers, and just launched what seems to be the first video embedding model: https://learn.mixpeek.com/vuse-v1-release/ https://le

hey HN, We're all NLP & CV engineers, and just launched what seems to be the first video embedding model: https://learn.mixpeek.com/vuse-v1-release/ https://learn.mixpeek.com/vuse-v1-release/ here are some example semantic queries: https://mixpeek.com/video?q=breaking+the+ice https://mixpeek.com/video?q=breaking+the+ice https://mixpeek.com/video?q=human+connection https://mixpeek.com/video?q=human+connection as you can see, these are not literal queries they're colloquialisms. which can't be done with simple transcription/object detection models or even CLIP. we deploy in our customers' VPC, and give you full control of the embeddings.