Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Clarification of performance #6

Open
frankcarey opened this issue Sep 20, 2012 · 2 comments
Open

Clarification of performance #6

frankcarey opened this issue Sep 20, 2012 · 2 comments

Comments

@frankcarey
Copy link

I have this working now! (images and data streaming to the computer, not the car part) I do see some things that look wrong, but I'm not sure since it looks like they may have been coded that way for a reason.

@dps - First,I really appreciate you putting this on github. Note that I'm not complaining here, just trying to document for those evaluating like I am. If you think these are actually bugs, I can create separate tickets. If not, maybe you can clarify here.

  1. only the top left corner of the pixels are outputting. My guess is that this is to reduce the size of the image or to be able to handle different camera sizes. The effect is that only a portion of the image you see on your phone's preview shows up in the Driver app on the computer. It might make sense to use the center of the image or resize it down.

  2. The image is greyscale - This was done on purpose ( decodeYUV420SPGrayscale() ) probably to save bandwidth or to make it easier for the neural net to deal with I'm guessing.

  3. The screen doesn't stay on and when the screen turns off, the images on the computer (Driver) stop updating.

@dps
Copy link
Owner

dps commented Sep 20, 2012

Hi Frank, comments inline

On Thu, Sep 20, 2012 at 11:57 AM, Frank Carey [email protected]:

I have this working now! (images and data streaming to the computer, not
the car part) I do see some things that look wrong, but I'm not sure since
it looks like they may have been coded that way for a reason.

Awesome!

@dps https://github.com/dps - First,I really appreciate you putting
this on github. Note that I'm not complaining here, just trying to document
for those evaluating like I am. If you think these are actually bugs, I can
create separate tickets. If not, maybe you can clarify here.

No problem - excited that you are using it!

  1. only the top left corner of the pixels are outputting. My guess is that
    this is to reduce the size of the image or to be able to handle different
    camera sizes. The effect is that only a portion of the image you see on
    your phone's preview shows up in the Driver app on the computer. It might
    make sense to use the center of the image or resize it down.

This is a bug. A full frame (but downsampled) image should be output.

  1. The image is greyscale - This was done on purpose (
    decodeYUV420SPGrayscale() ) probably to save bandwidth or to make it easier
    for the neural net to deal with I'm guessing.

Yes, on purpose to preserve bandwidth and feature size. The neural net
really shouldn't need color data,

  1. The screen doesn't stay on and when the screen turns off, the images on
    the computer (Driver) stop updating.

Not something I have run into, but the screen dimming timeout is
configureable per phone. Would be cool to take a wake lock to keep the
phone alive. Feel free to send me a pull request.


Reply to this email directly or view it on GitHubhttps://github.com//issues/6.

@prateekralhan
Copy link

@frankcarey and @dps Sir, Can we change the API level of the android app! As it says api level 14 i guess as per the prject.properties file! But the current updated api level is 27 and I am not able to downgrade it to 14. Please help here a bit sir or cna i make changes in the android manifest.xml file and run it accordingly!!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants