Jump to Content

Home Documentation

Python Package Log In

Home Documentation

Overview

What is Lightly?

Getting Started

Install Lightly
Set Up Your First Dataset
Run Your First Selection
Analyze the Selection Results
Export Filenames

Advanced

Videos as Input
Customize a Selection
- Selection Strategy Combination
- Active Learning Scorers
Train a Self-Supervised Model
Relevant Filenames
Datapool
Corruptness Check
Sequence Selection
Custom Embeddings
Configuration Options
Outputs
Automation

Tutorials

Active Learning Using YOLOv7 and Comma10k
Active Learning Using YOLOv7 for Waste Sorting
Active Learning for Driveable Area Segmentation Using Cityscapes
Active Learning for Transactions of Images
Improving YOLOv8 using Active Learning on Videos
Assertion-based Active Learning with YOLOv8
Active Learning for Sports Applications using Keypoint Detections

Providing Data to Lightly

Cloud Storage
Work with Metadata
Work with Predictions
- Prediction Format
- Lightly Pretagging

The Lightly Platform

Overview
Core Concepts
Analyze and Filter
Export
Metadata Configuration

Other Bits

Security
Hardware Recommendations
Known Issues & FAQ
Debugging
Contact Us
Assign Scheduled Runs to Specific Workers
Migration guides
- Migrating to Worker v2.4
- Migrating to Worker v2.5

Powered by

Core Concepts

There are many core concepts of understanding the Lightly platform. Head to one of the subpages to read about them!

Updated 4 months ago