Each region now has an appropriate placeholder image instead of a generic image and the get image button is a more obvious icon.
Reviewed-on: Georgi_Lab/ALVINN_f7#84
Closes: #77
If the user clocks on the image, any detections present at that location will be selecting in descending order of confidence.
Signed-off-by: Justin Georgi <justin.georgi@gmail.com>
Reviewed-on: Georgi_Lab/ALVINN_f7#78
Closes: #70
Adds a prelaoder animation to the results container while the image is loading and being processed for detections.
Signed-off-by: Justin Georgi <justin.georgi@gmail.com>
Reviewed-on: Georgi_Lab/ALVINN_f7#72
Moving panel to page element was resulting in opening errors after more than one page opened.
viewedAll computer property was returning errors before result data was fetched.
Signed-off-by: Justin Georgi <justin.georgi@gmail.com>
Reviewed-on: Georgi_Lab/ALVINN_f7#65
Closes: #48
This PR should allow the user to upload an image that has been taken with the camera and the associated detection information to a file server. These file can then be used to add data to ALVINN's models.
Current upload destination is set to a file drop folder in Georgi Lab Nextcloud, but eventually (maybe) should be changed to some dedicated ALVINN backend.
Also, prediction data is still not formatted in any specific way. It is just a JSON array of (top, left, bottom, right, label) objects.
Reviewed-on: Georgi_Lab/ALVINN_f7#56
Closes: #29
Fixes access to the camera in the android cordova build and maintains access when using the vite build.
Reviewed-on: Georgi_Lab/ALVINN_f7#45
Using inline svg in vue components will:
1. Make it possible to recolor svgs (for example for dark mode)
2. Be a better solution to the cordova resource location issue
Signed-off-by: Justin Georgi <justin.georgi@gmail.com>
Reviewed-on: Georgi_Lab/ALVINN_f7#32
This PR makes the detection page image loading a promise for better loading success and error handling.
Signed-off-by: Justin Georgi <justin.georgi@gmail.com>
Reviewed-on: Georgi_Lab/ALVINN_f7#25
Closes#22 and also fixes unlisted issue with the display of the placeholder image on the detect screen when in landscape mode.
Signed-off-by: Justin Georgi <justin.georgi@gmail.com>
Reviewed-on: Georgi_Lab/ALVINN_f7#24
Closes#3
Allows for configuration of model server settings (address and port) and can dynamically select the model based on the detection page selected.
Signed-off-by: Justin Georgi <justin.georgi@gmail.com>
Reviewed-on: Georgi_Lab/ALVINN_f7#16
Closes: #11
Only setting available at the moment are for pytorch server . But this is this easily expandable.
Signed-off-by: Justin Georgi <justin.georgi@gmail.com>
Reviewed-on: Georgi_Lab/ALVINN_f7#15
