OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Image & Video AI

Deepmosaics

DeepMosaics is a Python computer-vision application that uses semantic segmentation and image-to-image translation to automatically add or remove mosaic effects in images and videos.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Deepmosaics — OpenViST project profile in Image & Video AI.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use DeepMosaics when an image or video workflow specifically needs automatic mosaic addition or removal and you can provide an appropriate pre-trained model.

Who it is for

  • People editing still images to add or remove mosaic effects.
  • People processing videos with mosaic addition or removal.
  • Developers who want to run the project from source or train with their own dataset.

Documented uses

  • Remove a mosaic effect from an image or video.
  • Add a mosaic effect to an image or video.
  • Run the tool through a Windows GUI package or from source.
  • Train using a custom dataset, following the project's separate training guide.
Before you choose

Requirements & limits

The project documents a pre-built GUI package for Windows x86_64 and source use on Linux, Mac OS and Windows. It is open source under GPL-3.0.

  • The selected pre-trained model determines the intended effect; the project says models suit different effects.
  • The source workflow requires a mosaic-position model file in addition to the selected operation model.
  • The documented GPU option requires CUDA; processing time varies with computer performance.
  • The Windows GUI is listed for Windows x86_64, with Windows 10 described as preferable.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Run a documented mosaic operation from source

Prepare the documented source workflow and run an example to add or clean a mosaic.

Before you start

  • Linux, Mac OS or Windows
  • Python 3.6 or later
  • ffmpeg 3.4.6 and PyTorch 1.0 or later
  • OpenCV and torchvision, installed through the project requirements
  • CPU or NVIDIA GPU with CUDA CuDNN

Steps

  1. Clone the repository and change into its directory.

  2. Download pre-trained models and place them in ./pretrained_models.

  3. Install dependencies with pip install -r requirements.txt.

  4. Run deepmosaic.py with a media path and the chosen model path; use an add-face model to add mosaic or a clean-face model to remove it.

Expected outcome

The README says output media is saved in ./result.

Guide limitations

  • The examples use specific model files; the project says different pre-trained models suit different effects.
  • Processing time depends on computer performance; the GPU option requires CUDA.
  • If the mosaic-position model path error appears, the README directs users to check for mosaic_position.pth at the required path.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
15 October 2026
Original website title
HypoX64/DeepMosaics · Official homepage title · retained in its source language
How we review sourcesSuggest a correction