Home

Welcome to the LENS wiki!

Here you will find useful guides on how to use this repository and develop your own network for performing event based visual place recognition.

---

Table of Contents

• Train and evaluate your own custom datasets
  • Introduction
  • Preparing your data
  • Training your model
  • Evalutating your model
    • Ground truth evaluation
• Optimize your network hyperparameters
  • Introduction
  • Getting started
  • Hyperparameters
    • Good practices for hyperparameters
  • Setting up hyperparameters you want to fine tune
  • Other things to set
  • How we evaluate our model
• Deploy your model to a SynSense Speck2fDevKit
  • Introduction
  • Getting started
  • Evaluate your model on-chip using simulated event streams
  • Running the online inferencing model
    • Output from the online inferencing model
  • Create a dataset from the DevKit
  • Limitations

---

Contribute

LENS is a completely open source project comitted to improving the event based VPR community. If you have a suggestion or feature enhancement request, please submit it as an issue.

If you have added some functionality that you would like to contribute, please submit it as a pull request.