≈ the.bay.news

Tiled inference for upscaling images larger than VRAM: a working pattern

PyTorch Forums
Tiled inference for upscaling images larger than VRAM: a working pattern
Most upscaling models are trained on 256-512px crops but people want to run them on 4K photos. The model OOMs, so you tile. The naive tiling loop has two failure modes that are easy to miss. Failure mode 1: seam artifacts from independent normalization If each tile is normalized separately (per-tile mean/std, or per-tile min/max), the tiles come back at slightly different brightness and you get a visible grid. Normalize over the whole image first, then tile. Obvious in hindsight, but it bites ...

0 comments

Sign in to join the discussion — your thebay.events account works here.

No comments yet.