Epigos provides an end-to-end platform to annotate data, train computer vision AI models, deploy them seamlessly and host the models via API's.
For more details, visit epigos.ai.
The Epigos Python Package is a python wrapper around the core Epigos AI web application and REST API.
To install this package, please use Python 3.9
or higher.
To add epigos
to your project simply install with pip:
pip install epigos
Or with poetry
poetry add epigos
To make your first API call, you will need to signup at epigos.ai and create an API key for your workspace. Please contact our sales team for a demo.
import epigos
client = epigos.Epigos("api_key")
Manage project and upload dataset into your project using the Project ID
.
import epigos
from epigos.typings import BoxFormat
client = epigos.Epigos("api_key")
# load project
project = client.project("project_id")
# upload image with Pascal VOC annotation
record = project.upload(
"path/to/image.jpg",
annotation_path="path/to/image.xml",
box_format=BoxFormat.pascal_voc
)
print(record)
# upload image with COCO annotation
record = project.upload(
"path/to/image.jpg",
annotation_path="path/to/coco.json",
box_format=BoxFormat.coco
)
print(record)
# upload image with YOLO annotation
record = project.upload(
"path/to/image.jpg",
annotation_path="path/to/image.txt",
box_format=BoxFormat.yolo
)
print(record)
import epigos
client = epigos.Epigos("api_key")
# load project
project = client.project("project_id")
# upload COCO annotation dataset
records = project.upload_coco_dataset(
images_directory="path/to/dataset/train/images",
annotations_path="path/to/dataset/train/coco.json",
)
print(tuple(records))
# upload Pascal VOC annotation dataset
records = project.upload_pascal_voc_dataset(
images_directory="path/to/dataset/train/images",
annotations_directory="path/to/dataset/train/labels",
)
print(tuple(records))
# upload YOLO annotation dataset
records = project.upload_yolo_dataset(
images_directory="path/to/dataset/train/images",
annotations_directory="path/to/dataset/train/labels",
data_yaml_path="path/to/dataset/data.yaml",
)
print(tuple(records))
Make predictions with any of the models deployed in your workspace using the Model ID
.
import epigos
client = epigos.Epigos("api_key")
# load classification model
model = client.classification("model_id")
# make predictions
results = model.predict("path/to/your/image.jpg")
print(results.dict())
import epigos
client = epigos.Epigos("api_key")
# load object detection model
model = client.object_detection("model_id")
# make predictions
results = model.detect("path/to/your/image.jpg")
print(results.dict())
# visualize detections
results.show()
If you want to extend our Python library or if you find a bug, please open a PR!
Also be sure to test your code with the make
command at the root level directory.
Run tests:
make test
It’s important to write sensible commit messages to help the team move faster.
Please follow the commit guidelines
This project uses Semantic Versioning.
This project is published on PyPi
This library is released under the MIT License.