call-for-code / droneaid Goto Github PK

If your laptop is connected to the Tello Drone hotspot and you try to load the http://127.0.0.1:3000 webpage, the index.html tries unsuccessfully to load the remote CSS page
<link rel="stylesheet" href="https://codepen.io/ibmcodait/pen/gVMdwm.css">
and scripts

    <script src="https://cdn.jsdelivr.net/npm/@cloud-annotations/object-detection"></script>
    <script src="https://codepen.io/ibmcodait/pen/gVMdwm.js"></script>

There's some additional recursive dependences within those files too.

Since the laptop is not connected to the internet (at least over wireless), the rendered page is a mess. It would be better to move these dependencies local into this repo so that a Ctrl-R / page reload doesn't wreck the drone demo.

For now the workaround is to :

• Connect to your mobile smartphone hotspot
• npm start
• Open a browser tab to http://127.0.0.1:3000
• Power on the Tello Drone
• Connect to the Tello hotspot
• Don't reload the browser page.
• Fly the drone with object detection.

Is your feature request related to a problem? Please describe.
An (Anki) deck of flashcards would make learning and recalling the DroneAid symbols even easier

Describe the solution you'd like

• Deck
  • Card(s)
    • Image + Visual Description
    • Meaning

Describe alternatives you've considered

• js quiz
  • https://h5p.org/content-types-and-applications
  • jsfiddle w/ the source

Additional context

• https://en.wikipedia.org/wiki/Anki_(software)
• https://docs.ankiweb.net/getting-started.html
• https://docs.ankiweb.net/importing.html
• https://github.com/ankitects/anki-manual/blob/main/src/importing.md

Is your feature request related to a problem? Please describe.
The IBM Model Asset Exchange provides a way to reuse models instead of having to train them manually.

Describe the solution you'd like
The trained model materials need to be packaged and shared through MAX.

Describe alternatives you've considered
Training the model each time for each platform.

Additional context
This could be related to any other Dockerization steps.

Is your feature request related to a problem? Please describe.
Users currently have to download Homebrew, ffmpeg, Node.js and other tools. We should package this all in Docker. This also helps support Windows users.

Describe the solution you'd like
Create a Dockerfile that captures all the steps.

Describe alternatives you've considered
Platform-specific instructions

Additional context
The trained visual recognition model should also be shared through the IBM Model Asset Exchange as a Docker image.

Could a symbol be added to the table of symbols and the trained model for virus/plague/zombie apocalypse?

Describe the bug
The model detects one of the classified images and draws a bounding box but the counter on the web page does not increment

To Reproduce
Steps to reproduce the behavior:

1. View some of the symbol images.
2. Observe that the model draws bounding boxes.
3. The web page does not increment

Describe the bug
One of the NPM packages has a security issue. Committers can see it.

@pedrocruzio @vabarbosa @bourdakos1 @johnwalicki can you try to upgrade and test?

Is your feature request related to a problem? Please describe.
When objects have been detected, the interesting image frame should include GPS info so that the image can be plotted on the map

Describe the solution you'd like
See the map in #3

Describe the bug
The original Slack link expired after 30 days. We need a new one that does not expire.

when initiate the command cacli download it keep showing the syntax of the command is incorrect.

A written description of each symbol in the README table could be helpful for training and memory purposes.

I wrote this; please feel free to reuse none or any of it without attribution:

Each of the symbols are drawn within a triangle pointing up:

• Immediate Help Needed (orange; downward triangle \n SOS),
• Shelter Needed (cyan; like a guy standing in a tall pentagon without a floor),
• OK: No Help Needed (green; upward triangle \n OK),
• First Aid Kit Needed (yellow; briefcase with a first aid cross),
• Water Needed (blue; rain droplet),
• Area with Children in Need (lilac; baby looking thing with a diaper on),
• Food Needed (red; pan with wheat drawn above it),
• Area with Elderly in Need (purple; person with a cane)

Is your feature request related to a problem? Please describe.
The input stream to the visual recognition model should be source agnostic, whether from drones, civil aviation, or satellite feeds.

Describe the solution you'd like
Need to provide working code that shows this implementation.

Is your feature request related to a problem? Please describe.
We assume that the person in need will have a kit with the printed symbols available. We should improve the system to demonstrate how a person could hand-recreate the symbols, and in turn, make the recognition more sensitive to those symbols.

Is your feature request related to a problem? Please describe.
Beyond the current open source TensorFlow based implementation, we need to add steps for alternative visual recognition systems, including PowerAI and Watson Visual Recognition.

Describe the solution you'd like
DroneAid in general needs to have a flexible architecture that allows different components to implement its 4 core features; 1. Several device type inputs, 2. Several training systems, 3. Several realtime processing systems, and 4. Several downstream reporting/messaging outputs.

Additional context
The original system was based on PowerAI. We need to recover those steps if possible.

Per additional Linux Foundation guidance.

Is your feature request related to a problem? Please describe.
Need to replace the lost code in the PowerAI version for the plotting the located needs by type and coordinates on the map.

Describe the solution you'd like
This is the fourth and final component needed for an end-to-end version of DroneAid.

Additional context
The initial version of DroneAid had this dashboard as part of the solution. It was lost when the PowerAI version implementation was deleted.