Using Transfer Learning to solve the simulator-to-real problem in the Duckietown environment

Introduction

Transfer Learning

Applied methods

Domain Randomization

Observations from the standard (upper row) and the domain randomized (lower row) Duckietown environment

Image Thresholding

The original and the thresholded images from both the simulated and real-world domains

Visual Domain Adaptation using UNIT networks

UNIT network-based Visual Domain Adaptation
The quality of the UNIT network-based image-to-image translation: sim-to-real (upper images), real-to-sim (lower images)

Results

The results of the Transfer Learning experiments

References:

--

--

--

Deep Learning and AI solutions from Budapest University of Technology and Economics. http://smartlab.tmit.bme.hu/

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

R-CNN Object Detection and Semantic Segmentation

Introduction to Overfitting, Underfitting and Data Mismatch in Building Machine Learning Systems

Teaching a Machine to Trade Stocks like Warren Buffett, Part I

Gradient Descent for Multivariable Regression in Python

Getting to Know Hao Zhang

How to Forecast Website Traffic

Graph Theory + Implementation

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
SmartLab AI

SmartLab AI

Deep Learning and AI solutions from Budapest University of Technology and Economics. http://smartlab.tmit.bme.hu/

More from Medium

Review: VPGNet — A Multi-Task Network for Autonomous Driving

Organizing a Mess With Computer Vision

GFPGAN — Towards Real-World Blind Face Restoration with Generative Facial Prior

Computer Vision — Making the machines see