What challenges did Camera Futura face before AI implementation?

Photographers spend a lot of time manually culling low-quality images and organizing photos before post-processing. The workflow is difficult when shots are very similar, when photographers need to choose between RAW and JPG images, and when technical qualities like sharpness and white balance must be judged.

What was the impact of the AI solution?

Camera Futura says ML.NET enabled it to build, train, and deploy a production-level machine learning model quickly. The founder said delivering the first model production-ready would have needed several more months without ML.NET. The product opened new opportunities for the startup by enabling features that would have been difficult with multiple technologies.

Source

MicrosoftProductionEvidence: Medium55/100

Camera Futura Automates Photo Culling and Organizing with ML.NET

Use case typeComputer vision inspectionUpdated Jan 1, 2021

Camera Futura, based in Geneva, Switzerland, built Futura Photo, a desktop application that automates photo culling and organizing before post-processing. The product uses ML. NET image classification and clustering models trained on company images to assess technical criteria such as sharpness and white balance and to group similar images for selection. The application is a WPF desktop app on . NET Framework, with models trained locally and then deployed with the desktop software.

Organization: Camera Futura
Industry: Tech & Comms
Location: Switzerland
Published: January 2021

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Camera Futura
Provider: Microsoft
Maturity: Production
Linked source: Microsoft .NET Customer Story

NET Framework, with models trained locally and then deployed with the desktop software

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 2 of 2

1Computer vision inspection
2Workflow automation

Camera Futura built Futura Photo as a desktop application that lets photographers configure rules, upload images, review results, and move photos to the right folder.
The solution uses three models built with ML.NET Model Builder plus a clustering model built with the ML.NET API.
The models analyze images for features such as sharpness, white balance, and similar-image grouping to automate culling and organizing.

Technologies

ML.NET ML.NET Model Builder Visual Studio WPF .NET Framework Azure

Camera Futura says ML.NET enabled it to build, train, and deploy a production-level machine learning model quickly.
The founder said delivering the first model production-ready would have needed several more months without ML.NET.
The product opened new opportunities for the startup by enabling features that would have been difficult with multiple technologies.

Architecture

Futura Photo is a WPF desktop application on .NET Framework 4.6.1. It trains on company images locally, with models deployed alongside the desktop application. Three production models were created with ML.NET Model Builder image classification, and one clustering model was built with the ML.NET API.

Sources & evidence1

Evidence: Medium55/100Evidence strength