The project aims to create media analysis plug-ins for the Autopsy - Digital Forensics platform. With the use of open-source machine learning engines and technologies, we have been able to create a suite of plug-ins on the Python programming language that integrates and extends the functionality of Autopsy.
- Object Detection and Identification for still images.
- Optional Facial Recognition.
- Based on ImageAI
- Object detection and identification on video files.
- Based on ImageAI
- Audio to text transcription.
- Based on DeepSpeech
.
├── DeepSpeech/ # AudioToText plugin
├── VideoAI/ # Video Analyser plugin
│ └── models/ # imageAI pretrained model
├── imageAI/ # Image Analyser plugin
│ └── models/ # imageAI pretrained model
├── sample/ # Sample test data
└── README.md
- Microsoft Windows 10
- Autopsy
4.19.1
- Java
1.8
64-bit - Python
3.7
Versions and dependencies are listed within the installation guide
- Tensorflow
2.4.0
- ImageAI
2.1.6
- face_recognition
1.3.0
- DeepSpeech
0.9.3
- ImageAI v1.0
- DeepSpeech v0.9.3
Due to potential conflicts in dependency versions it is highly recommended that you install the following files in the provided sequence.
-
The basics
- Ensure you have Python
3.7
installed! Download here - Python Dependencies
pip install --upgrade pip
pip install wheel
- Ensure you have Python
-
Image Analyser / Video Analyser
- Python Dependencies
pip install tensorflow==2.4.0 keras==2.4.3 numpy==1.19.3 pillow==7.0.0 scipy==1.4.1 h5py==2.10.0 matplotlib==3.3.2 opencv-python keras-resnet==0.2.0 imageai
- Others
- The ImageAI model has been pre-bundled
- If you wish to replace the bundled model you can do so at the plugin sub-directory
./imageAI/models/yolo-tiny.h5
and./VideoAI/models/yolo-tiny.h5
- Python Dependencies
-
Facial Recognition for Image Analyser
- Python Dependencies
pip install face_recognition
- If the install fails on Windows when installing
dlib
, please install
- Python Dependencies
-
AudioToText
- Python Dependencies
pip install deepspeech pydub webrtcvad
- Others
- FFMPEG to support conversion from
.mp3
to.wav
for processing - DeepSpeech Model to be downloaded and selected when prompted
- FFMPEG to support conversion from
- Python Dependencies
-
Either download via
- The release bundle Latest Release
- Or any other method e.g
git
-
Move the following folders into the Autopsy Python plugin directory
DeepSpeech
VideoAI
imageAI
Plugin directory on Windows is
%appdata%\autopsy\python_modules\
Your directory tree should now look like
python_modules\ ├── DeepSpeech\ # AudioToText plugin ├── VideoAI\ # Video Analyser plugin └── imageAI\ # Image Analyser plugin
-
Ensure DeepSpeech Models have been downloaded as per Prerequisites - 4. AudioToText
-
Launch Autopsy!