deepdataspace

DeepDataSpace

The Go-To Choice for CV Data Visualization, Annotation, and Model Analysis

DeepDataSpace (DDS) is an innovative open-source tool tailored for handling datasets with three primary functions: data visualization, annotation, and model analysis. This tool is designed to be user-friendly and efficient, offering features that make managing datasets more interactive and insightful.

Key Features

• Interactive Dataset Visualization and Exploration: DDS allows for in-depth visualization of datasets, enabling users to explore and understand their data better.
• Intelligent Annotation with Collaborative Workflow: It provides intelligent annotation tools which facilitate collaboration among team members, streamlining the annotation process.
• Efficient Model Management and Performance Analysis: Users can manage models effectively and analyze their performance to ensure optimal outcomes.

Installation

Prerequisites

To get started with DDS, you'll need Python versions ranging from 3.8 to 3.10. The tool is compatible with several operating systems and platforms, including:

• Mac OS (x86/x64 and arm64)
• Windows 10 (x86/x64)
• Ubuntu LTS from version 18.04 (x86/x64)
• Docker Compose (x86/x64 and arm64)

Installing from PyPI

You can easily install DDS using Python’s package manager, pip. First, make sure pip is upgraded, and then install the package using the following commands:

python3 -m pip install pip --upgrade
python3 -m pip install deepdataspace

Quick Start

After installation, the dds command will be available for initiating the tool. You can start DDS quickly with:

dds --quickstart

This command will set up and run DDS on your local machine. You can then access the tool via a web browser at http://127.0.0.1:8765, where you can view sample datasets.

Alternative Installation Methods

From Source Code

To install DDS from the source code:

1. Clone the repository from GitHub.
2. Set up the Node environment if needed.
3. Compile frontend files and prepare them for integration into the Python package.
4. Finally, install using pip.

Using Docker Compose

Another method involves using Docker Compose, which requires cloning the repository, setting up a dataset directory, choosing a port, and running DDS with Docker. This process also enables you to access DDS through a local web address.

Documentation

For more detailed guidance, DDS has comprehensive documentation available here, including quick start guides, tutorials, API references, and more.

Uninstallation

To uninstall DDS, use the following commands depending on your installation method:

• For PyPI or Source Code Installation: Use pip to uninstall and delete runtime files.

  pip uninstall deepdataspace
  rm -rf ~/.deepdataspace/* # Be cautious as this deletes all imported datasets
  

• For Docker Installation: Stop the container, remove the Docker image, and volumes.

  docker stop dds
  docker rmi deepdataspace/dds
  docker volume remove dds-runtime # Be cautious as this deletes all imported datasets
  

License

DeepDataSpace is licensed under the Apache 2.0 License, allowing it to be used freely under the terms specified.

By offering robust features for visualization, annotation, and model analysis, DDS serves as a comprehensive tool for working with computer vision datasets. Whether you are a researcher, developer, or data scientist, this tool can significantly enhance your dataset management capabilities.