Awards winning music video

Rachel Claudio:
    "Water is the commonest symbol for the unconscious.
    The lake in the valley is the unconscious, which lies, as it were, underneath consciousness, so that it is often referred to as the “subconscious,” usually with the pejorative connotation of an inferior consciousness” - Carl Jung”
    The more we learn about the subconscious, the more we understand to what degree our conscious decisions are influenced by it.
    Attachment theory demonstrates how the dynamics of our personal relationships are set up between the ages of 0-3 - in the “programming years”. Humans are an interdependent species. We are wired for attachment as a means of survival. Human relationships are intended to be set up as gratifying & energising so that we can survive our vulnerabilities & draw upon each other’s strengths as needed. But when this is not the case for a small human, the mesolimbic system in the brain begins to rewire or “miswire” itself so that relationships begin to be synonymous with emotional pain & exhaustion.
    This means that even if we are fortunate enough to stumble upon the person we ultimately want to be with, the programming is likely to determine whether or not that relationship is successful & healthy or destructive & toxic - despite the greatest intents & will of the people involved. The subconscious, in it’s linking of what is known to what is safe, strives to “protect” us with the familiar. Even if the familiar is “negative attachment” made manifest in pain & sabotage. In the mind of someone with disorganised, anxious or insecure attachment, the subconscious will incessantly interject itself to interrupt flow & faith. 

Martins Dabolins:
    I Will shortly tell you how I come up with the video.
    When Rachel first time sends me a preview of the song it just melted within me. We had some discussion about how she sees and feels about video - we both were on the same boat, deep underwater, dark looking scene, far distant body, and super slow. Footage from the shooting was different, fast, bright, and close. I did some experiments and got that classic techniques will not work out or it will take forever to achieve the right effect. I started with trimming out best shoots and slowing them down from 50fps to 600-800 fps. non of the pro video software couldn't achieve good results. I was thinking maybe Ai Deep Networks can solve the problem better than algorithms. In an experiment, I come up with different solutions and pre-trained models, but there wasn't any underwater pre-trained model and to train one by my self I didn't have underwater super slow motion footage with persona inside. I chose a model from a car of-road rally. Flying dust and stones were very similar to bubbles and movement in the scene, that this solution was way superior to others. The next problem was zoom out from close up and there again I use Ai Deep Networks to segment the body out of the background. For segmentation, I use the yolact++ library that can detect and segment not only humans but animals and different casual objects but not humans underwater, so I chose to detect any human and animal, raised very high detection error level that Ai interpolates and sees different things that not there as our mind do. I was told to draw square masks around newly detected objects and I use that squares as a mask to segment out humans from the scene. The last thing left was to draw back the background. I didn't want to take some random underwater scene and put it under I chose to stretch pixels from square edges out that they overlap with each other and fill up empty spaces. I like that effect, it feels like blurred glass, like a mind You think You know, but not sure, free space for imagination. Every time I look at that video it fully grabs my attention it just surprises me how accurate it represents the human condition. Fragmental, asynchronous - synchronous, fogy, and clear.