Visual Explanations#

Build
Backend
Frontend

Components#

Predictions are provided by a MobileNetV2 model pre-trained on ImageNet and fine-tuned on task-specific samples
Backend microservice built with FastAPI
Frontend created with VueJS utilizing the Advanced Cropper component

Environment Variables#

Variable	Description	Default Value
`SERVICE_NAME`	The name that the app uses as its identifier, e.g. when logging or emitting traces	`"inspection-service"`
`ROOT_PATH`	FastAPI root path (doc)	`""`
`PATH_PREFIX`	FastAPI router prefix (doc)	`""`
`DEFAULT_EXPLAINER`	The explanation generator to use if a request does not speficy one	`"lime"`
`LOG_INPUT`	Save input images to disk	`False`
`LOG_PATH`	Path to store the logged images	`./log`

Running the service locally (with Docker)#

To launch a standalone instance of the service:

cd visual-inspection
docker-compose up

Running the service locally (without Docker)#

During development, it can be desirable to launch the service directly and not as a Docker container. For instance, this allows to get rapid feedback on changes to the backend code.

cd visual-inspection
./build_frontend.sh
cd inspection-backend
uvicorn inspection.main:app

Log input images#

For development purposes, the service can be configured to write the raw input images to disk. This setting is controlled through the environment variable LOG_INPUT:

export LOG_INPUT=1

By default, the images are collected in a sub-folder ./log. To change where the images are saved to, set the environment variable LOG_PATH, e.g.:

export LOG_PATH=/home/myusername/log