CNIT 581 ATR – RoboPatrol Final Blog Post

Introduction This week we presented the final demo of our project. We have learned a lot working on this project and have grown as researchers and advocates of assistive technology along the process. Before we go, here is a recap of our project once more. Project Overview Navigating an indoor environment requires awareness of continually…

Update 12/3

APP Design: Due to the nuances of our project space, we spent significant time in the past few weeks brainstorming and ideating on the potential scope, interaction modalities, and behavior for the app. We particularly paid great attention to the interactional aspects of the app because it is important that we develop a product that evokes…

Project Update 11/19/2021

This week we have trained 10,000 episodes of YOLO. Here are some examples of recognition results. Next we will add more data and train our model to 30,000 episodes. And we will test the dynamic recognition in real time video.

Project Update 12/11/2021

This week, we completed the labelling work of the objects to be recognized in robot vision system. Our labelling process included over 500 samples. Next week, we will train our CNN model YOLO based on the labeled samples. Images of some of the labelled items are below. Gamification and Crowdsourcing The success of our system…

Loading…

Something went wrong. Please refresh the page and/or try again.

About us

Hello, we are the student of Purdue University! This blog is to post the update of our team project (Robo-Patrol) for CNIT 581 ATR (Fall 2021).
~ Ike, Prakash and Ruiqi.

Subscribe to Our Blog

Get new content delivered directly to your inbox.