Error in image segmentation tflite model running on android demo app?

TensorFlow General Discussion
1

Hi ,
I trained a tflite for image segmentation with MobileNetV3+DeeplabV3+PascalVOC, check out the model and compare with deepLab_v3, the diff input quantization, how to change the input quant to fit the input of Android tflite demo app, thanks a lot.

image
image1415×556 27.9 KB

image
image1718×603 32.9 KB

3

Attach pb tflite convert script

import os
import numpy as np
import tensorflow as tf

print(tf.__version__)

MODEL_FILE = "frozen_inference_graph_257.pb"

# Load the Tensorflow model
converter = tf.compat.v1.lite.TFLiteConverter.from_frozen_graph(
    graph_def_file = MODEL_FILE,
    input_arrays = ['sub_2'], # For the Xception model it needs to be `sub_7`, for MobileNet it would be `sub_2`
    output_arrays = ['ResizeBilinear_2'],
    input_shapes={'sub_2':[1,257,257,3]}
)

# Default Optimization is choosen
converter.optimizations = [tf.lite.Optimize.DEFAULT]

# Convert to TFLite Model
tflite_model = converter.convert()

# Save the model as tflite format.
tflite_path = "deeplabv3_mnv2_custom_257.tflite"

tflite_model_size = open(tflite_path, 'wb').write(tflite_model)
4

Procedure:

git clone --depth 1 https://github.com/tensorflow/models.git

curl -sc /tmp/cookie "https://drive.google.com/uc?export=download&id=1rATNHizJdVHnaJtt-hW9MOgjxoaajzdh" > /dev/null
CODE="$(awk '/_warning_/ {print $NF}' /tmp/cookie)"
curl -Lb /tmp/cookie "https://drive.google.com/uc?export=download&confirm=${CODE}&id=1rATNHizJdVHnaJtt-hW9MOgjxoaajzdh" -o pascal_voc_seg/VOCtrainval_11-May-2012.tar


sed -i -e "s/python .\/remove_gt_colormap.py/python3 .\/remove_gt_colormap.py/g" -i -e "s/python .\/build_voc2012_data.py/python3 .\/build_voc2012_data.py/g" download_and_convert_voc2012.sh

sh download_and_convert_voc2012.sh

mkdir -p deeplab/datasets/pascal_voc_seg/exp/train_on_train_set/train
mkdir -p deeplab/datasets/pascal_voc_seg/exp/train_on_train_set/eval
mkdir -p deeplab/datasets/pascal_voc_seg/exp/train_on_train_set/vis

export PATH_TO_TRAIN_DIR=${HOME}/deeplab/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_train_set/train
export PATH_TO_DATASET=${HOME}/deeplab/models/research/deeplab/datasets/pascal_voc_seg/tfrecord
export PYTHONPATH=${HOME}/deeplab/models/research:${HOME}/deeplab/models/research/deeplab:${HOME}/deeplab/models/research/slim:${PYTHONPATH}

python deeplab/train.py \
    --logtostderr \
    --training_number_of_steps=500000 \
    --train_split="train" \
    --model_variant="mobilenet_v3_small_seg" \
    --decoder_output_stride=16 \
    --train_crop_size="513,513" \
    --train_batch_size=8 \
    --dataset="pascal_voc_seg" \
    --save_interval_secs=300 \
    --save_summaries_secs=300 \
    --save_summaries_images=True \
    --log_steps=100 \
    --train_logdir=${PATH_TO_TRAIN_DIR} \
    --dataset_dir=${PATH_TO_DATASET}

python deeplab/export_model.py --model_variant="mobilenet_v3_small_seg" --crop_size=257 --crop_size=257 --checkpoint_path=${PATH_TO_TRAIN_DIR}/model.ckpt-1871 --export_path=./deeplabv3_mnv2_pascal_trainval/frozen_inference_graph_257.pb

`

5

examples/lite/examples/image_segmentation/android
git clone https://github.com/tensorflow/examples