Unable to load asset

[malder08]

I'm using the recognizeBinaryImage and imageToByteListFloat32 function from the example in conjunction with the imagePicker function where I pass in the selected image file from another screen. When running the recognizeBinaryImage function, an error is thrown at this line:
var imageBytes = (await rootBundle.load(image.path)).buffer;
The error I receive is
Unable to load asset: /storage/emulated/0/Android/data/com.example.tensorflow_image_classification_gci/files/Pictures/20191205_133618.jpg
Another thoughts on what the problem could be?

[mtkgeek]

I'm having this issue too. Any help will be appreciated

[malder08]

@mtkgeek Are you having this issue when using the default model? If you are, I was unfortunately never able to get it working with the default model but when I switched over to the ssd mobilenet model it was working fine. I'm not entirely sure what was causing it not to work with the default model but try using a different model and see if that works for you

[mtkgeek]

@malder08 I'm using an automl vision pretrained model and i think it's quantized. I guess tflite still has issues working with quantized models. Thanks anyways.

[Ehtasha]

@malder08 @mtkgeek Did you guys resolve this problem ?
I'm also facing this problem.

[malder08]

@Ehtasha If you are using the default model, then I never got around to fixing the problem. Unless you have to, I would just use another model like the ssd mobilenet model.

[jherkyesse]

@malder08
this is my workaround. hope it would help you

import 'dart:io' as Io;
var imageBytes = Io.File(image.path).readAsBytesSync();