Autonomous Movement from Simultaneous Localization and Mapping

Overview

Autonomous Movement from Simultaneous Localization and Mapping

About us

Built by a group of Clarkson University students with the help from Professor Masudul Imtiaz and his Lab Resources.

Micheal Caracciolo           - Sophomore, ECE Department
Owen Casciotti               - Senior, ECE Department
Chris Lloyd                  - Senior, ECE Department
Ernesto Sola-Thomas          - Freshman, ECE Department
Matthew Weaver               - Sophomore, ECE Department
Kyle Bielby                  - Senior, ECE Department
Md Abdul Baset Sarker        - Graduate Student, ECE Department
Tipu Sultan                  - Graduate Student, ME Department
Masudul Imtiaz               - Professor, Clarkson University ECE Department

This project began in January 2021 and was finished May 5th 2021.

Synopsis

Presenting the development of a Simultaneous Localization and Mapping (SLAM) based Autonomous Navigation system.

Supported Devices:

Jetson AGX
Jetson Nano

Hardware:

Wheelchair
Jetson Development board
Any Arduino
Development Computer to install Jetson Jetpack SDK (For AGX)
One Intel Realsense D415
One Motor controller ()
2 12V Batteries For Motors
2 12V Lipo Batteries for Jetson

Software:

Tensorflow Version: 2.3.1

OpenVSLAM

We will need to install a few different Python 3.8 packages. We recommend using Conda environments as then you will not have to compile a few packages. However, some packages are not available in Conda, for those just install via pip while inside of the appropriate Conda env.

csv
heapq
Jetson.GPIO (Can only be installed on Jetson)
keyboard
matplotlib
msgpack
numpy
scipy (Greater than 1.5.0)
signal
websockets

Initial Setup

OpenVSLAM, Official Documentation

Webserver, Not needed unless want to interface with phone

  • Move the www folder into your /var directory in your root file system.
  • Open up python server files and insert your static IP of your Jetson
  • Run python server.py

Note: There is some example data and maps in the csv format. This format is required to correctly transmit maps/paths to the device that is listening to the server.

Android Phone, APK here

  • Insert the IP wanting to connect to, in this instance, the static IP of the Jetson
  • Build the Java app to your Android Phone

Note: This can only be used if the Webserver is set up and the server.py is on. We recommend to have it be turned on via startup. We do not have this implemented in our current code, but can be easily added. If you plan on using a Android Phone for a Map/Path/End point interface, you will need to edit some lines in /src/main.py and add to send_location.py. This is all untested code currently.

Source Code, ensure you're in the right Conda Environment

  • To use your own map/.msg file from OpenVSLAM, you will need to put it in the /data folder. There are a few options with this, you can either use the raw .msg file which our MapFileUnpacker.py will take care of, or you can create a csv format of 0 and 1's in the format of a map. 0 being unoccupied and 1 being occupied in the Occupancy Grid Map. For even easier storage, you could run MapFileUnpacker.py and have it extract the keyframes into a csv, which then you can use for OLD_main.py or main.py. We recommend to use the map file you created which is in the form of .msg.
  • You can either use OLD_main.py or main.py. OLD_main.py can be ran without having to run the motors on the connected Jetson. This is helpful for debugging and testing before you decide to implement the map onto a Jetson. main.py will ONLY work on a Jetson as it will call JetsonMotorInterface.py which contains Jetson.GPIO libraries which can only be installed on a Jetson.
  • If the Android Phone is set up, you will need to edit main.py to send the start position via send_location.py to the webserver. You will also need to uncomment a few lines so that the current map is sent to the /var/www/html filepath. Then, the phone should be able to send back a end value which calls def main with that end value. Otherwise, def main will run with a predefined end value in code.
  • To set up the pinout, you will need to first build arduino_motor_ctrl.ino onto the Arduino that is connected to the motor controller. You can use virtually any pins on the Arduino, depending on what Arduino you use. Set these pins in the .ino file. Next, we want to set the pins on the Jetson that output the data to the Arduino pins. Set these pins in JetsonMotorInterface.py. Be careful not to use any I2C or USART pins as these cannot be configured as GPIO Output.

Note: To properly run main.py without any issues, it is recommended to follow this so that you do not need to run Sudo for any of the /src files. If you were to run Sudo, you would have a bunch of different libraries and it will not run properly. If you get an Illegal Instruction error, please try to create a Conda environment to run these scripts.

Note: We are using a Sabertooth 2x32 Dual 32A Motor Driver to drive our dual Wheelchair motors. The Arduino also gets it's power from the Motor Driver, but do not connect it there while it is connected to the computer for building.

A few things to be weary of, in the main.py, since we are not using the Localization from VSLAM, we are simulating the created map into a path. This path will run differently depending on how accurate it is and the speed of your motors. We recommend you to scale your room to your map, so you will want to section out your map in code and have a timing ratio to ensure it moves the right distance of "Occupancy Grid Map spaces". This is explained better in the code.

The Reinforcement Learning files inside of /src/RL are purely experimental and do work for training. However, due to time constraints, they have not been polished enough to work with our design. They are published here for any future use as they are completely made open-source.

A simple program for training and testing vit

Vit This is a simple program for training and testing vit. Key requirements: torch, torchvision and timm. Dataset I put 5 categories of the cub classi

xiezhenyu 2 Oct 11, 2022
A PoC Corporation Relationship Knowledge Graph System on top of Nebula Graph.

Corp-Rel is a PoC of Corpartion Relationship Knowledge Graph System. It's built on top of the Open Source Graph Database: Nebula Graph with a dataset

Wey Gu 20 Dec 11, 2022
A denoising autoencoder + adversarial losses and attention mechanisms for face swapping.

faceswap-GAN Adding Adversarial loss and perceptual loss (VGGface) to deepfakes'(reddit user) auto-encoder architecture. Updates Date Update 2018-08-2

3.2k Dec 30, 2022
Learning where to learn - Gradient sparsity in meta and continual learning

Learning where to learn - Gradient sparsity in meta and continual learning In this paper, we investigate gradient sparsity found by MAML in various co

Johannes Oswald 28 Dec 09, 2022
Multi-Stage Progressive Image Restoration

Multi-Stage Progressive Image Restoration Syed Waqas Zamir, Aditya Arora, Salman Khan, Munawar Hayat, Fahad Shahbaz Khan, Ming-Hsuan Yang, and Ling Sh

Syed Waqas Zamir 859 Dec 22, 2022
Deep High-Resolution Representation Learning for Human Pose Estimation

Deep High-Resolution Representation Learning for Human Pose Estimation (accepted to CVPR2019) News If you are interested in internship or research pos

HRNet 167 Dec 27, 2022
《Fst Lerning of Temporl Action Proposl vi Dense Boundry Genertor》(AAAI 2020)

Update 2020.03.13: Release tensorflow-version and pytorch-version DBG complete code. 2019.11.12: Release tensorflow-version DBG inference code. 2019.1

Tencent 338 Dec 16, 2022
EASY - Ensemble Augmented-Shot Y-shaped Learning: State-Of-The-Art Few-Shot Classification with Simple Ingredients.

EASY - Ensemble Augmented-Shot Y-shaped Learning: State-Of-The-Art Few-Shot Classification with Simple Ingredients. This repository is the official im

Yassir BENDOU 57 Dec 26, 2022
DeLiGAN - This project is an implementation of the Generative Adversarial Network

This project is an implementation of the Generative Adversarial Network proposed in our CVPR 2017 paper - DeLiGAN : Generative Adversarial Net

Video Analytics Lab -- IISc 110 Sep 13, 2022
Implicit Graph Neural Networks

Implicit Graph Neural Networks This repository is the official PyTorch implementation of "Implicit Graph Neural Networks". Fangda Gu*, Heng Chang*, We

Heng Chang 48 Nov 29, 2022
NeuralDiff: Segmenting 3D objects that move in egocentric videos

NeuralDiff: Segmenting 3D objects that move in egocentric videos Project Page | Paper + Supplementary | Video About This repository contains the offic

Vadim Tschernezki 14 Dec 05, 2022
A library for low-memory inferencing in PyTorch.

Pylomin Pylomin (PYtorch LOw-Memory INference) is a library for low-memory inferencing in PyTorch. Installation ... Usage For example, the following c

3 Oct 26, 2022
A program that uses computer vision to detect hand gestures, used for controlling movie players.

HandGestureDetection This program uses a Haar Cascade algorithm to detect the presence of your hand, and then passes it on to a self-created and self-

2 Nov 22, 2022
CoReD: Generalizing Fake Media Detection with Continual Representation using Distillation (ACMMM'21 Oral Paper)

CoReD: Generalizing Fake Media Detection with Continual Representation using Distillation (ACMMM'21 Oral Paper) (Accepted for oral presentation at ACM

Minha Kim 1 Nov 12, 2021
Deep GPs built on top of TensorFlow/Keras and GPflow

GPflux Documentation | Tutorials | API reference | Slack What does GPflux do? GPflux is a toolbox dedicated to Deep Gaussian processes (DGP), the hier

Secondmind Labs 107 Nov 02, 2022
Pytorch implementation of Deep Recursive Residual Network for Super Resolution (DRRN)

DRRN-pytorch This is an unofficial implementation of "Deep Recursive Residual Network for Super Resolution (DRRN)", CVPR 2017 in Pytorch. [Paper] You

yun_yang 192 Dec 12, 2022
PyTorch 1.0 inference in C++ on Windows10 platforms

Serving PyTorch Models in C++ on Windows10 platforms How to use Prepare Data examples/data/train/ - 0 - 1 . . . - n examples/data/test/

Henson 88 Oct 15, 2022
LeafSnap replicated using deep neural networks to test accuracy compared to traditional computer vision methods.

Deep-Leafsnap Convolutional Neural Networks have become largely popular in image tasks such as image classification recently largely due to to Krizhev

Sujith Vishwajith 48 Nov 27, 2022
Bridging Composite and Real: Towards End-to-end Deep Image Matting

Bridging Composite and Real: Towards End-to-end Deep Image Matting Please note that the official repository of the paper Bridging Composite and Real:

Jizhizi_Li 30 Oct 31, 2022
Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation (CVPR 2021)

Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation (CVPR 2021, official Pytorch implementatio

Microsoft 247 Dec 25, 2022