Case Study

Adding 3D and Point-Cloud Annotation to a Widely Used CV Tool

The client's tool gained advanced 3D capabilities while keeping the ease of use that made it popular in the first place.

Industry

AI, Data & MLOps

Time Frame

2 Months

Services We offer

Computer Vision

Frontend / 3D

Python

Data Tooling

About the client

A global leader in semiconductor technology.

Problem Statement

An annotation tool that excelled at 2D needed first-class 3D support for autonomous systems, robotics and complex datasets, without disrupting the experience its users already relied on.

The Challenges

3D without disruption

Adding point-cloud annotation had to preserve the 2D experience users already depended on.

Formats that matter

Support for Velodyne and KITTI made the tool useful for real autonomous-systems datasets.

Approach

We extended the tool rather than replacing it:

3D rendering and point-cloud annotation
Seamless switching between 2D and 3D views
Referential image mapping for context and precision
Drag-and-drop workflow and support for Velodyne and KITTI formats

The Solution

Users annotate and navigate point clouds, switch between 2D and 3D to visualize complex data, and lean on multiple viewpoints and shortcuts to work faster, built with Three.js on a Django and Postgres backend, containerized for deployment.

Read more case studies

An End-to-End Model Preparation and Validation Pipeline

Learn More

Cutting Claim Turnaround by 70% with Document AI and RPA

Learn More

One Dashboard Across 8+ DevOps Tools

Learn More

Get Detailed Case Study

Adding 3D and Point-Cloud Annotation to a Widely Used CV Tool

Integrating 3D rendering and point-cloud annotation into an established 2D annotation tool, so users move seamlessly between 2D and 3D for autonomous-systems and robotics datasets.

Tráiganos el caso extremo que su nube no puede alcanzar. Vamos a diseñarlo.