automatic-behavior-analysis 0.0.21

Program complex for automated behavior analysis

AutomaticBehaviorAnalysis

Installation

pip

Client (in cache mode) To work with the cache, you need to place the video and cache file in the same folder.

Package installation:

pip install --user automatic-behavior-analysis

Package update:

pip install --upgrade --user automatic-behavior-analysis

Client launch:

aba-client

ATTENTION! For correct operation, the package path should not contain Cyrillic characters.

Requirements

• Python 3
• TensorFlow
• Tensorflow Object Detection API
• Keras
• numpy
• PIL
• appdirs
• openCV
• websocket
• imageio

Client2

• pip install PyQt5 opencv-python appdirs requests matplotlib scipy pandas

Build a container with a server

Development container

There are no files and models in the development container in order to minify the image and increase the usability.

The container is collected by the following command:

# For container with GPU support 
docker build -t registry.gitlab.com/digiratory/automatic-behavior-analysis/dev-gpu .
# For container with CPU support only 
docker build  -f Dockerfile.dev.cpu -t registry.gitlab.com/digiratory/automatic-behavior-analysis/dev-cpu .

There's no need to rebuild the container for work and you can pick up the assembled from the repository using the following commands:

docker login registry.gitlab.com
# To run a container with GPU support 
docker run --runtime=nvidia -it -v <workspace>:/home/user/ -p 1217:1217 registry.gitlab.com/digiratory/automatic-behavior-analysis/dev-gpu:latest

# For container with CPU support only 
docker run -it -v <workspace>:/home/user/ -p 1217:1217 registry.gitlab.com/digiratory/automatic-behavior-analysis/dev-cpu:latest

For CPU-only:

docker login registry.gitlab.com
docker run -it -v <workspace>:/home/user/ -p 1217:1217 registry.gitlab.com/digiratory/automatic-behavior-analysis/dev-cpu:latest

If the machine does not have gpu, then the key --runtime=nvidia no need to specify.

ATTENTION! The development container does not include source code, models, and more.

Launch applications

Launch order:

1. Server part
2. Client side

Shutdown Procedure:

1. Client side
2. Server part

Server

To start the server side, you must run ./server.py

python3 ./server.py

Team Arguments: [ip= ] — Ip wiretap address (Default 172.0.0.1 ) [port= ] — Server Port (Default 1217)

Client part (GUI)

To start the client side, you must run ./client_gui.py

python3 ./client_gui.py

Team Arguments: [ip= ] — Ip server address (Default 172.0.0.1 ) [port= ] — Server Port (Default 1217)