This repository has been archived by the owner on Apr 7, 2020. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 2
Weekly Report 05 (05.13. ~ 05.17.)
NXXR edited this page May 17, 2019
·
1 revision
- Online Courses as introduction and refresher for machine learning
- TurtleBot2 Introduction
- Project Scope
- Plans for Week 06
- Machine Learning Online Course by Stanford University & coursera.org started and ~20% done
- Deep Learning Onramp by MatLab started and ~40% done
- preliminary read of the TurtleBot Tutorials/Documentation
Stage is a 2D multi-robot simulator- Gazebo is a 3D environment simulating camera input
- behaves similar to a 3D editor
- models are loaded into a 3D environment
- Corridor & Junction Detection
- Detecting the Room Layout and Junctions in Corridors, create trajectories through junctions according to target waypoint.
- End-to-End Driving Model
- Training of an End-to-End Driving model to move along unobstructed corridors to a given waypoint.
- Route Planning
- Implement route planning algorithms to navigate mobile robot through complex environments along waypoints.
Optional Goals 4. Object Detection
- Train neural network to detect static and mobile objects (obstacles) and modify or set waypoints to avoid obstacles.
- Floor Type Detection
- Train neural network to detect changes in surface texture and detect dirt/stains. Mark position of detected surface as target waypoint.
Distinguishing features to other projects:
Feature | HCU Project | End-to-End Navigation | End-to-End Learning of Driving Models |
---|---|---|---|
Single Sensor System | O | X | X |
Route Planning | O | junction trajectories | external system (GPS) |
Obstacle Detection | O | X | O |
- Formalize Scope and Survey Results
- Continue Online Courses and Tutorials
- Introduction TurtleBot & deeper look into Robot Simulation