Saltar al contenido
KaiConvert
IA localPrevisto

Upscale an image

Planned as a local AI tool: enlarge an image with detail that plain interpolation cannot produce.

Todavía no disponible

Esta herramienta está en la hoja de ruta. La página existe para documentar el flujo, los formatos y las herramientas relacionadas, pero no se procesa nada ni se produce ningún resultado.

IA local
Ver las herramientas que ya funcionan

This page has not been translated yet and is shown in English. It is excluded from search indexing for this language until it has been.

Conventional resampling can make an image bigger but not sharper — it has no information to add, so it spreads the existing pixels more thinly. Super-resolution models predict plausible detail instead, which is why an AI upscale looks different in kind from a bicubic one.

Super-resolution models are small enough to be a realistic local-AI candidate, which is why this is planned as Tier B rather than a cloud call. It ships when on-device inference is dependable.

Qué obtienes

  • 2× and 4× targets

    The standard super-resolution factors, chosen per image.

  • On-device inference planned

    A small enough model to run in a browser with WebGPU or WebAssembly.

  • Tiled processing

    Large images processed in tiles to keep memory use bounded.

Cómo funciona

  1. 1

    Capability check

    WebGPU or WebAssembly availability is confirmed before the feature is offered.

  2. 2

    Tiled inference

    The image is processed in overlapping tiles inside a Web Worker.

  3. 3

    Reassemble

    Tiles are recombined onto the canvas and the result enters the history.

Especificaciones

  • This tool is not implemented yet. Nothing is processed, no image is uploaded, and no result is produced. The page documents the intended workflow and is excluded from search indexing until the feature actually works.
  • Planned as Tier B (local AI): ONNX Runtime Web with tiled inference.

Por qué usar KaiConvert

  • Sin subidas en las herramientas locales
  • Funciona sin conexión una vez cargada la página
  • La IA en la nube siempre se indica antes de empezar

Preguntas frecuentes

Herramientas relacionadas