Galliot now hosts all Neuralet’s content and expertise gained in three years of work and completing high-quality applications, mainly in Computer Vision and Deep Learning. More Info

Galliot Face Anonymizer with DeepStream

This repository contains the codebase related to our article on creating a face anonymizer application with Nvidia DeepStream. galliot have published three other articles which walk you through the features of Nvidia DeepStream and its python Bindings. Part 1 Part 2 Part 3

Prerequisites

1. X86 device with a Nvidia GPU
2. CUDA 11.1+
3. Docker and Nvidia Docker Toolkit

Getting Started

1. Clone The repository:

git clone https://github.com/galliot-us/deepstream-example.git

cd deepstream-example

1. build the docker image:

docker build -f face_annonymizer_ds_51_x86.Dockerfile -t "galliot/face_anonymizer_ds:x86" .

2. Run the docker image:

docker run -it  --runtime nvidia --gpus all -v "$PWD/":/repo galliot/face_anonymizer_ds:x86

3. Run the prepare_models.bash to download the object detectors and build the required modules:

bash ./prepare_models.bash

4. run one of runner scripts based on the chosen object detector:

For SSD MobileNet V2 Detector:

python3 deepstream_face_anonymizer_ssd.py --config config_infer_primary_ssd_mobilenet.txt --label_path labels.txt  --input_video [PATH OF THE INPUT VIDEO] --out_dir [PATH OF THE OUTPUT VIDEO FILE]

For YOLO V3 SPP Detector:

python3 deepstream_face_anonymizer_yolo.py --config config_infer_primary_yolo_mobilenet.txt --label_path labels.txt  --input_video [PATH OF THE INPUT VIDEO] --out_dir [PATH OF THE OUTPUT VIDEO FILE]