2022-02-03

 Date

Feb 3, 2022

 Participants

  • @Mika Shaw

  • @Sahil Kabir

  • @Aaryan Shroff

  • @Jerry Bi (Deactivated)

  • @Vivek Bhardwaj

  • @Michael Aizenshtadt

  • @Mihir Gupta

  • @Denny (Unlicensed)

  • @Krishna Kalavadia

 Discussion topics

Task

Notes

To Do:

Task

Notes

To Do:

Task 1 ML training dataset

  • Done

Nothing

Task2 collect ML testing set

  • Not much has been done

  • try other ways of collecting data for testing set

  • research

Task3 model comparison

  • started

  • google drive + docs file created with research notes

  • looked into mAP

  • look into model differences from top view and stanford drone

Task 4 DeepSORT

  • Mock up of detect function running

  • reformat to take in arguments

  • continuing getting familiar with warg repo

Task 5 explosive pickup brainstorming

  • Trying different ideas - contour detection

  • Try building + implementing simple edge detection

Task 6 yolov5 model incremental training

  • Training on Stanford drone 99% (using best.pt)

  • Optimization on top view 96%

  • share with task 3

  • continue training

Task 7 Video interface of the groud station

  • Done

  • displayed video

  • Connect video player to ground station

Task 8 Geolocation

  • Talked to Ryan + got solid understanding

  • continue geining a better understanding + teesting

Task 9 repo maintainence & unit test

  • No updates + ongoing as no updates have been made

  • unit test that needs to be done (not too sure)

Task 10 Integration test

 

  • Will test modules given by Ray

Task 12 map labelling

  • looking into google maps API

  • creating a sample

  • customize for comp needs

 Decisions