• Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    AI
    Announcement

    A photo gallery demo trains image alignment as users drag rectangles

    The creator says the JavaScript gallery uses backpropagation in real time and includes a slider for its loss parameter.

    2 Sources, 32m ago, first seen 32m ago

    TLDR

    The creator distinguishes trained neural nets from hand-written weights that match a program exactly. They describe a JavaScript photo gallery that trains as rectangles are dragged, making photos in another view align. In a separate demo, they say mostly traditional code was transformed into WebGPU matrices to search 100,000 JSONs in real time, with regex, parsing, highlighting and matching forming one exact neural net.

    Combined views

    —

    2 Sources, first seen 32m ago

    — likes— comments— saves— reposts

    Combined views

    —

    2 Sources, first seen 32m ago

    — likes— comments— saves— reposts
    Image: @_chenglou on X

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Featured Source

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    2 Sources

    Cheng Lou@_chenglouWhat happens in we unfreeze some of these weights and actually train them? Well you can apply existing neural net techniques to it, losses and backprops and all, and now the UI becomes a bit more “learned” Here’s a photo gallery where on the left I edit the rectangles and the right side gives me a new view where the photos learn to align themselves, training in real time as you drag (because the param count is so tiny). This demo done fully in JavaScript, with JS looping through them numbers and backpropping. Stupid but fun! I’ve exposed the loss param slider you can play around with32m
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI

    2 Sources

    Cheng Lou@_chenglouWhat happens in we unfreeze some of these weights and actually train them? Well you can apply existing neural net techniques to it, losses and backprops and all, and now the UI becomes a bit more “learned” Here’s a photo gallery where on the left I edit the rectangles and the right side gives me a new view where the photos learn to align themselves, training in real time as you drag (because the param count is so tiny). This demo done fully in JavaScript, with JS looping through them numbers and backpropping. Stupid but fun! I’ve exposed the loss param slider you can play around with32m
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet