Skip to content
This repository has been archived by the owner on May 27, 2021. It is now read-only.

Latest commit

 

History

History
112 lines (106 loc) · 4.99 KB

progression.md

File metadata and controls

112 lines (106 loc) · 4.99 KB
Raw

Progression Note

  • 22/11/2020
    • already train solid model for specific task(3 cap type + 1 unknows)
    • Try to modify detect.py in order to work with mygui.py (kinda hard)
    • very lazy implementation (do what minimum)
  • 26/11/2020
    • understanding Threading in YOLOv5
    • try to use threading in order to save detected image to some specific location
    • lazy implementation (do what it should)
  • 27/12/2020
    • look into detect.py again so i would modify it to be used with mygui.py
    • new feature proposal (count,live video feed in 1 Giant GUI)
    • exploring Machine learning and essentail library to complete this project
    • i have no idea how to do next step
  • 25/1/2021
    • disable saving image after finding 'unknown' object (for performance later // will fix later)
    • can count now (lazy implmentation)
    • will try to squeeze all feature in GUI (still don't know what to do )
  • 14/3/2021
    • i filped it all !!!!! reconstuction all my work into 3 seperate unit
    • my part is only handling client side so..... no more re-train yea !
  • 16/3/2021
    • finished reconstucting GUI .now seperating into 2 tab .
    • awaiting API endpoint from server side
    • still can't integrate CV2 live video output into myGUI
    • try using API with my own server (can send/get result of unknown image)
  • 20/3/2021
    • VDO streaming within window application is available !!!
    • try to manipulate detect.py to be used inside mygui.py as a function
  • 21/3/2021
    • success ! now using mygui_detect.py for directly using YOLOv5 for this project
    • successfully seperate preparation code and detection code so i could work around with loading time
    • using thread for sure !
    • YOLOv5 implementation is almost completed.
  • 22/3/2021
    • MyGUI.py is good to go ! . no more using subprocess to call Yolov5
    • Server YOLOv5 is up and running !
  • 31/3/2021
    • Cleaning Repo from clean_work branch
    • try to delete non-essential file out of this project coz' it too messy!
  • 3/4/2021
    • Cleaning completed !
    • now client+server code can work together in single Repo
  • 6/4/2021
    • can send image to detect at yolo-server
    • time to process result is pretty good (0.2 sec)
  • 19/4/2021

    • fix server file because why not ? it a part of my job

    • using token is available now

    • need to integrate registration and login into myGUI

    • first implement GUI for login and register . Quick and simple

    • put black box behind count indicator

  • 20/4/2021
    • new UI
    • new function (yet to implement)
  • 21/4/2021
    • registration ok
    • constant unknown data to be sent to YOLO-server ok (poor performance)
    • will do image "swiping" later
  • 24/4/2021

    • can swap model now

    • waiting for model downloading from YOLO-server when it finished

    • daily count is possible

    • not sure how to collect new image to be re-train tho..

    • prototyping "collect new image" feature . not sure if it working correctly...

    • new folder path defined again.

  • 26/4/2021
    • can swap model (from server) now
    • can send 50 image for re-training
    • what do now ?
  • 29/4/2021
    • overhaul UI (so customer could use my program easily)
  • 1/5/2021
    • detection is possible
    • applicaiton structure feel like it too complexed.... but it work fine
  • 2/5/2021
    • from now on , update will be appear down below
  • 4/5/2021
    • this project could have better optimization and better error handler... i'll do later AFTER getting a grade from this project first
  • 20/5/2021
    • after deadline work so..... it gonna be rough
    • "Smart-swap" feature is implemented
    • hey, if it work , it ain't wrong
  • 26/5/2021
    • last edit before presentation
    • good luck .

OVERHAUL UI Progression (v1.7+)

Status Feature Status Feature
✅ preload YOLO model ✅ detect and stop
✅ login to YOLO-server ✅ folder & file for GUI
✅ list all Model (Local&Sever) ✅ download server model
✅ Swap a new model ✅ register to YOLO-server
✅ send unknown ✅ swiping in view mode
✅ one click capture ✅ send unknown
✅ file handling ✅ test connection
✅ get unknown result ✅ view mode
✅ capture mode ✅ Friendly-UI
⌛ Warning and exception handler ⌛ better optimization
⌛ Cross-platform ⌛ executable program

New File Location

  • BASE_DIR\unknown : for saving unknown image to be sent to server
  • BASE_DIR\unknown\raw : for saving raw image to be sent to server
  • BASE_DIR\gui_data : for saving file to be used in building GUI
  • BASE_DIR\mine : for saving local model file