braindotai/Watermark-Removal-Pytorch

Health v2 · maintenance only

Adoption not part of the score

1148 stars · 166 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

A PyTorch implementation of watermark removal based on the Deep Image Prior paper, using a CNN generator's structure to restore images without training data. It includes a simple API wrapper and supports CUDA and MPS devices.

Use cases

• remove watermarks from images
• restore images with known watermark masks
• remove artefacts from photos without training a model
• run deep image prior image restoration in pytorch
• clean watermarked images when the watermark overlay is available

When to choose

• you have the watermark (or mask) matching the image's position and scale
• you want training-free image restoration with realistic results
• you want a lightweight model with fast inference on GPU or Apple MPS

When to avoid

• you need to remove watermarks without any knowledge of the watermark or its position
• you need batch processing of thousands of images at production scale
• you need a general-purpose inpainting or photo editing suite

Facets

library · maturity maintenance

image-processing machine-learning deep-learning computer-vision image-processing deep-learning python cross-platform watermark-removal deep-image-prior pytorch image-restoration inpainting gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="braindotai/Watermark-Removal-Pytorch")