Hey, I'm mostly focusing on an Android alternative to ARKit human segmentation for masking out the background while leaving a hand in view.
When using ARKit human segmentation, the results here are great as you can see, however I have struggled to find something that can produce similar results on Android
What would you advise to someone using this repo with the Deeplab example to improve the segmentation of hands?
Hello. The application you have provided works perfectly awesome with the provided tflite file. We tried the same approach replacing the model, with newer version of deeplab from Tensorflow model zoo(https://github.com/tensorflow/models/blob/master/research/deeplab/g3doc/model_zoo.md). The inference obtained is, there is difference in the input nodes and output nodes, between their pb and your tflite. Any suggestions or ideas on how to overcome this. Due to the difference, we are getting an issue in the loadmodel function.. Actual error is "ByteBuffer is not a valid flatbuffer model". If you need, we can provide you a Log file for the same. And also do we have to make any changes in the asset variables, if we use the recent tflite?
I generate a model.tflite based on MobileNetV2_BiSeNet and run on the python code,it's right.
I change the params of TF_OD_API_MODEL_FILE , TF_OD_API_NUM_CLASS=2(because my segament bg and person),TF_OD_API_LABELS_FILE(bg and person)
I just make some change above all.
then I generate an apk ,install on my phone,but it can not run right.
can you give some suggstions that i can think about what is wrong possibly.
sir excuse me if my issue doesn't relate to this repo, but I will appreciate any help.
i used deeplab model on real-time ios application but what i got is very slow capture, i don't know what is the problem is it the deeplab model(.mlmodel size=8mo) or ios device(ios=12.0) or....??
any idea how can I improve my application
I trained deeplab(mobilenetV2) by quantization-aware training method.
And I exported quantized pb file. ( Cropped by here )
And converted from pb file to tflite file.
So I made my quantized tflite file.
until Crop Size 321, segmentation result is good. ( Green Mask is Overlayed )
But Over Crop Size 321, segmentation is not good ( Green Mask is not Overlayed )
Could you teach me why this problem is happened?
Sorry for bad English and Thank you.