Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,37 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+assets/logo.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00001.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00002.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00003.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00004.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00005.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00006.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00007.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00008.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00009.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00010.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00011.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00012.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00013.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00014.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00015.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00016.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00017.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00018.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00019.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00020.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00021.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00022.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00023.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00024.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00025.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00026.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00027.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00028.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00029.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00030.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00031.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00032.webp filter=lfs diff=lfs merge=lfs -text
+mock_pictures/00033.webp filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,228 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[codz]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#   Usually these files are written by a python script from a template
+#   before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py.cover
+*.lcov
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+# Pipfile.lock
+
+# UV
+#   Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+# uv.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+# poetry.lock
+# poetry.toml
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#   pdm recommends including project-wide configuration in pdm.toml, but excluding .pdm-python.
+#   https://pdm-project.org/en/latest/usage/project/#working-with-version-control
+# pdm.lock
+# pdm.toml
+.pdm-python
+.pdm-build/
+
+# pixi
+#   Similar to Pipfile.lock, it is generally recommended to include pixi.lock in version control.
+# pixi.lock
+#   Pixi creates a virtual environment in the .pixi directory, just like venv module creates one
+#   in the .venv directory. It is recommended not to include this directory in version control.
+.pixi/*
+!.pixi/config.toml
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule*
+celerybeat.pid
+
+# Redis
+*.rdb
+*.aof
+*.pid
+
+# RabbitMQ
+mnesia/
+rabbitmq/
+rabbitmq-data/
+
+# ActiveMQ
+activemq-data/
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.envrc
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#   JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#   be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#   and can be added to the global gitignore or merged into this file.  For a more nuclear
+#   option (not recommended) you can uncomment the following to ignore the entire idea folder.
+.idea/
+
+# Abstra
+#   Abstra is an AI-powered process automation framework.
+#   Ignore directories containing user credentials, local state, and settings.
+#   Learn more at https://abstra.io/docs
+.abstra/
+
+# Visual Studio Code
+#   Visual Studio Code specific template is maintained in a separate VisualStudioCode.gitignore 
+#   that can be found at https://github.com/github/gitignore/blob/main/Global/VisualStudioCode.gitignore
+#   and can be added to the global gitignore or merged into this file. However, if you prefer, 
+#   you could uncomment the following to ignore the entire vscode folder
+.vscode/
+# Temporary file for partial code execution
+tempCodeRunnerFile.py
+
+# Ruff stuff:
+.ruff_cache/
+
+# PyPI configuration file
+.pypirc
+
+# Marimo
+marimo/_static/
+marimo/_lsp/
+__marimo__/
+
+# Streamlit
+.streamlit/secrets.toml
+
+# ai
+
+.claude/
+.gemini/
+.qwen/
+.pi/
+

README.md

@@ -0,0 +1,95 @@
+---
+tags:
+  - watermark-detection
+  - image-embedding
+  - yolo
+  - dinov2
+  - computer-vision
+  - pytorch
+license: mit
+pipeline_tag: object-detection
+---
+
+<p align="center">
+  <img src="assets/logo.webp" width="320" alt="MimiLens logo">
+</p>
+
+# MimiLens-Models
+
+This repository contains models for **[MimiLens](https://github.com/MinthyBasis/mimilens)** — a toolkit for automatic detection of consistent artifacts and watermarks in image datasets.
+
+
+High-quality training data is essential for generative models, because persistent artifacts, watermarks and repeated visual patterns in training images tend to leak into generated outputs. Manual inspection of large datasets is expensive and impractical, and even large multimodal models are costly and imperfect — especially when modern generators can produce text, speech bubbles and other objects that must be preserved.
+
+MimiLens solves this by looking for **consistency** across images instead of matching fixed templates. The pipeline runs in two stages:
+
+1. A fast YOLO-based detector proposes candidate regions across all images.
+2. A custom DINO-based embedder vectorizes these regions, and HDBSCAN clustering finds groups of visually similar detections.
+
+This separates repeating artifacts, watermarks, logos, and signatures from irregular, one-off objects.
+
+The models are lightweight and work on CPU, while a GPU significantly speeds up large-scale processing.
+
+# Models
+
+| File | Size | Format | Purpose |
+| --- | --- | --- | --- |
+| `models/detect_y26l_v1.2.pt` | ~50.7 MB | Ultralytics/PyTorch checkpoint | Detector |
+| `models/embedder_v3.0.safetensors` | ~84.7 MB | `safetensors` | Region embedder |
+
+## Detection
+
+The detection model is based on **YOLOv26** (`YOLO26l`). It was chosen as a recent architecture that offers a good balance between accuracy and inference speed.
+
+The training dataset combined real labeled images (the main part) with procedurally generated samples. The detector recognizes 5 classes:
+
+1. Signatures
+2. Watermarks
+3. Regular text
+4. Speech bubbles
+5. Comic / manga effects
+
+## Embeddings
+
+The embedder is built on top of **DINOv2-Small**. Its output embedding size was reduced to 128 dimensions for efficient clustering, and additional output layers perform the required transformations.
+
+The off-the-shelf DINOv2 is not directly suitable for this task because it focuses on foreground/background semantics rather than fine object-level shapes.  To obtain the required behavior, this model was modified and trained on a combined dataset (50% procedural generation from templates, 50% real data) using supervised-contrastive loss with cross-batch memory. This achieved high accuracy on the embedding task while keeping the model small.
+
+# Test data
+
+`mock_pictures/` contains stock photos and generated images with synthetic watermarks and signatures. They can be used to check how MimiLens behaves on typical cases.
+
+# Download
+
+Download the files manually and copy them into the `./models` folder of the MimiLens application, or use the Hugging Face CLI:
+
+```bash
+huggingface-cli download Minthy/MimiLens-Models --local-dir .
+```
+
+# Intended use
+
+These models are designed to be used through the MimiLens application and its pipeline (detection → embedding → clustering → review). For installation, GUI/CLI usage and configuration, please refer to the [MimiLens repository](https://github.com/MinthyBasis/mimilens).
+
+# Feedback
+
+Join the [Discord server](https://discord.gg/ZXHENAhqE9).
+
+Your examples can help the project improve.
+
+# Acknowledgments
+
+Drac, NeuroSenko, Sv1.
+
+# Donations
+
+BTC: `bc1qwv83ggq8rvv07uk6dv4njs0j3yygj3aax4wg6c`
+
+ETH/USDT(e): `0x04C8a749F49aE8a56CB84cF0C99CD9E92eDB17db`
+
+XMR: `47F7JAyKP8tMBtzwxpoZsUVB8wzg2VrbtDKBice9FAS1FikbHEXXPof4PAb42CQ5ch8p8Hs4RvJuzPHDtaVSdQzD6ZbA5TZ`
+
+# License
+
+The MimiLens project code, documentation and the `embedder_v3.0.safetensors` model are licensed under the **MIT License**.
+The `detect_y26l_v1.2.pt` detection model is based on YOLOv26 and is therefore licensed under **AGPL-3.0**.

models/detect_y26l_v1.2.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8326c1aa821439ae24bcfbbf8643281d397c04d58bdc0e17819b7fceef7a25e5
+size 53142117

models/embedder_v3.0.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22685ff791e3f6d52189b2f4aa10a90f702e40ec29c60d6312d77cc77650e9a8
+size 88782608