forgeryscope 0.0.3

Tools for the competition

Forgeryscope

Forgeryscope is a Python package for scientific image forgery detection. It contains tools for panel detection, embedding, and image/panel matching.

This is a simplified and refactored version of my winning solution for the Kaggle Scientific Image Forgery Detection competition.

For the full competition approach and design notes, see SOLUTION.md.

Features

• Panel detection with YOLO-based extractors
• Image embeddings with PyTorch/timm checkpoints
• Panel matching with LightGlue, SIFT, ALIKED, and geometry helpers
• On-demand model download from GitHub Releases

Linux Installation

Create and activate a virtual environment:

python3 -m venv .venv
source .venv/bin/activate
python -m pip install --upgrade pip

Install Forgeryscope from PyPI:

pip install forgeryscope

The package is available on PyPI: https://pypi.org/project/forgeryscope/

To install the latest version directly from GitHub:

pip install git+https://github.com/vlad3996/forgeryscope.git

If you work from a local clone:

git clone https://github.com/vlad3996/forgeryscope.git
cd forgeryscope
pip install -e .

Model Weights

Model weights are not stored in git and are not bundled into the Python package. They are downloaded on first use and cached locally.

Upload these files as assets to a GitHub Release, for example release tag models-v1:

• yolo_panel_extractor.pt
• yolo_lane_extractor.pt
• aliked_wblot.pth
• wblot_duplicate_embedder.ckpt
• wblot_overlap_embedder.ckpt
• wblot_lane_embedder.ckpt
• micro_overlap_embedder.ckpt

Then set the model URL in your Linux shell:

export FORGERYSCOPE_MODEL_BASE_URL="https://github.com/vlad3996/forgeryscope/releases/download/models-v1"

The first run downloads each requested file to:

~/.cache/forgeryscope

To use a different cache directory:

export FORGERYSCOPE_CACHE_DIR="/path/to/cache"

Quick Start

from ultralytics import YOLO

from forgeryscope import (
    Embedder,
    LightGlueOverlap,
    PanelExtractor,
    get_model_path,
    load_aliked_wblot_weights,
)

device = "cuda"

panel_extractor = PanelExtractor(
    weights_path="yolo_panel_extractor",
    device=device,
    conf_threshold=0.7,
    iou_threshold=0.4,
    verbose=False,
)
panel_extractor.EXCLUDED_LABELS = {"Graphs", "Flow Cytometry", "Body Imaging"}

lane_extractor = YOLO(get_model_path("yolo_lane_extractor"))

matcher_micro = LightGlueOverlap(
    max_keypoints=4096,
    matcher_features="sift",
    device=device,
    depth_confidence=0.9,
    width_confidence=0.9,
    verbose=False,
)

matcher_blot = LightGlueOverlap(
    max_keypoints=512,
    matcher_features="aliked",
    device=device,
    depth_confidence=-1,
    width_confidence=-1,
    estimator_method="MAGSAC",
    reprojThreshold=3.0,
    estimator_confidence=0.9999,
    estimator_maxIters=5000,
    estimator_refineIters=10,
    verbose=False,
)
load_aliked_wblot_weights(matcher_blot)

wblot_duplicate_embedder = Embedder("wblot_duplicate_embedder", device=device, verbose=False)
wblot_overlap_embedder = Embedder("wblot_overlap_embedder", device=device, verbose=False)
wblot_lane_embedder = Embedder("wblot_lane_embedder", device=device, verbose=False)
micro_overlap_embedder = Embedder("micro_overlap_embedder", device=device, verbose=False)

You can also pass a local checkpoint path instead of a model name:

embedder = Embedder("/path/to/wblot_duplicate_embedder.ckpt", device="cuda", verbose=False)

Available Model Names

• yolo_panel_extractor
• yolo_lane_extractor
• aliked_wblot
• wblot_duplicate_embedder
• wblot_overlap_embedder
• wblot_lane_embedder
• micro_overlap_embedder

Publishing Notes

Keep checkpoints/ ignored by git. Large .pt, .pth, and .ckpt files should live in GitHub Releases, not in normal repository history.

When a checkpoint changes:

1. Upload the new file to a new GitHub Release tag.
2. Update FORGERYSCOPE_MODEL_BASE_URL to the new release URL.
3. Update the SHA256 value in forgeryscope/model_zoo.py.

Requirements

• Python >= 3.8
• PyTorch >= 1.9.0
• torchvision >= 0.10.0
• OpenCV >= 4.5.0
• Dependencies listed in pyproject.toml

License

This project is licensed under the MIT License. See LICENSE.

Author

Uladzislau Leketush ([e-mail])