Skip to content Skip to footer
0 items - $0.00 0

Arbitrary-Scale Super-Resolution with Neural Heat Fields by 0x12A

Arbitrary-Scale Super-Resolution with Neural Heat Fields by 0x12A

Arbitrary-Scale Super-Resolution with Neural Heat Fields by 0x12A

8 Comments

  • Post Author
    jiggawatts
    Posted March 15, 2025 at 12:53 pm

    The learned frequency banks reminded me of a notion I had: Instead of learning upscaling or image generation in pixel space, why not reuse the decades of effort that has gone into lossy image compression by generating output in a psychovisually optimal space?

    Perhaps frequency space (discrete cosine transform) with a perceptually uniform color space like UCS. This would allow models to be optimised so that they spend more of their compute budget outputting detail that's relevant to human vision. Color spaces that split brightness from chroma would allow increased contrast detail and lower color detail. This is basically what JPG does.

  • Post Author
    Hizonner
    Posted March 15, 2025 at 1:43 pm

    Where are the ground truth images?

  • Post Author
    WhitneyLand
    Posted March 15, 2025 at 1:49 pm

    Seems like a nice result but wouldn’t have hurt for them to give a few performance benchmarks. I understand that the point of the paper was a quality improvement, but it’s always nice to reference a baseline for practicality.

  • Post Author
    flerchin
    Posted March 15, 2025 at 2:07 pm

    I'd like to see the results in something like Wing Commander Privateer.

  • Post Author
    adhoc32
    Posted March 15, 2025 at 2:29 pm

    Instead of training on vast amounts of arbitrary data that may lead to hallucinations, wouldn't it be better to train on high-resolution images of the specific subject we want to upscale? For example, using high-resolution modern photos of a building to enhance an old photo of the same building, or using a family album of a person to upscale an old image of that person. Does such an approach exist?

  • Post Author
    seanalltogether
    Posted March 15, 2025 at 2:43 pm

    I would love to see this kind of work applied to old movies from the 30s and 40s like the Marx Brothers.

  • Post Author
    nthingtohide
    Posted March 15, 2025 at 3:08 pm

    DLSS will benefit greatly from research in this area. DLSS 4 uses transformers.

    DLSS 3 vs DLSS 4 (Transformer)

    https://www.youtube.com/watch?v=CMBpGbUCgm4

  • Post Author
    flufluflufluffy
    Posted March 15, 2025 at 3:33 pm

    Was anyone else expecting an infinitely zoomable pictures from that title? I am disappoint

Leave a comment

In the Shadows of Innovation”

© 2025 HackTech.info. All Rights Reserved.

Sign Up to Our Newsletter

Be the first to know the latest updates

Whoops, you're not connected to Mailchimp. You need to enter a valid Mailchimp API key.