Convertion error on faster_rcnn_resnet50_keras

[GustavoLMourao]

Hello everyone.

I'm trying to converts the model faster_rcnn_resnet50_keras (which has the pipeline attached). However, I have got this error message:

'''

Model Optimizer arguments: Common parameters: - Path to the Input Model: None - Path for generated

Stopped shape/value propagation at "StatefulPartitionedCall/Preprocessor/ResizeToRange/cond" node

'''

I'm using DL Workbench (last version).

model {
faster_rcnn {
num_classes: 3
image_resizer {
keep_aspect_ratio_resizer {
min_dimension: 640
max_dimension: 640
pad_to_max_dimension: true
}
}
feature_extractor {
type: "faster_rcnn_resnet50_keras"
batch_norm_trainable: true
}
first_stage_anchor_generator {
grid_anchor_generator {
height_stride: 16
width_stride: 16
scales: 0.25
scales: 0.5
scales: 1.0
scales: 2.0
aspect_ratios: 0.5
aspect_ratios: 1.0
aspect_ratios: 2.0
}
}
first_stage_box_predictor_conv_hyperparams {
op: CONV
regularizer {
l2_regularizer {
weight: 0.0
}
}
initializer {
truncated_normal_initializer {
stddev: 0.01
}
}
}
first_stage_nms_score_threshold: 0.0
first_stage_nms_iou_threshold: 0.7
first_stage_max_proposals: 300
first_stage_localization_loss_weight: 2.0
first_stage_objectness_loss_weight: 1.0
initial_crop_size: 14
maxpool_kernel_size: 2
maxpool_stride: 2
second_stage_box_predictor {
mask_rcnn_box_predictor {
fc_hyperparams {
op: FC
regularizer {
l2_regularizer {
weight: 0.0
}
}
initializer {
variance_scaling_initializer {
factor: 1.0
uniform: true
mode: FAN_AVG
}
}
}
use_dropout: false
dropout_keep_probability: 1.0
share_box_across_classes: true
}
}
second_stage_post_processing {
batch_non_max_suppression {
score_threshold: 0.0
iou_threshold: 0.6
max_detections_per_class: 100
max_total_detections: 300
}
score_converter: SOFTMAX
}
second_stage_localization_loss_weight: 2.0
second_stage_classification_loss_weight: 1.0
use_matmul_crop_and_resize: true
clip_anchors_to_image: true
use_matmul_gather_in_matcher: true
use_static_balanced_label_sampler: true
use_static_shapes: true
}
}
train_config {
batch_size: 8
data_augmentation_options {
random_horizontal_flip {
}
}
sync_replicas: true
optimizer {
momentum_optimizer {
learning_rate {
cosine_decay_learning_rate {
learning_rate_base: 0.04
total_steps: 25000
warmup_learning_rate: 0.013333
warmup_steps: 2000
}
}
momentum_optimizer_value: 0.9
}
use_moving_average: false
}
fine_tune_checkpoint: "pre-trained-models/faster_rcnn_resnet50_v1_640x640_coco17_tpu-8/checkpoint/ckpt-0"
num_steps: 1000
startup_delay_steps: 0.0
replicas_to_aggregate: 8
max_number_of_boxes: 100
unpad_groundtruth_tensors: false
fine_tune_checkpoint_type: "detection"
use_bfloat16: false
fine_tune_checkpoint_version: V2
}
train_input_reader {
label_map_path: "annotations/label_map.pbtxt"
tf_record_input_reader {
input_path: "annotations/train.record"
}
}
eval_config {
metrics_set: "coco_detection_metrics"
use_moving_averages: false
batch_size: 1
}
eval_input_reader {
label_map_path: "annotations/label_map.pbtxt"
shuffle: false
num_epochs: 1
tf_record_input_reader {
input_path: "annotations/test.record"
}
}

[Syamimi_Intel]

Hi Gustavo Mourao,

Thank you for reaching out. We are currently investigating this issue and we like to know more information about your model.

What is the source of your model? And what is the version of TensorFlow that you used?

It would be ideal if you could share the model with us and how you convert the model, so we can reproduce and investigate the issue.

And, could you provide the log file? For that click the settings icon (gear at the right top corner) and download the log file by clicking the 'Download Log' button.

Regards,

Syamimi Fauzi

[VladimirG]

Hi Gustavo,

It seems like you have not selected a transformation config during the model conversion process. For the majority of models available in the Tensorflow Model Zoo, an appropriate transformation config is required to properly perform the layer replacement.

The DL Workbench provides a list of pre-defined configs for use during the conversion process. Just select "Use transformation config" option on the conversion page and select one from the table. For your model in particular "faster_rcnn_support_api_v2.0" should do the trick.

Regards,
Vladimir Golubenko

[Syamimi_Intel]

Hi Gustavo Mourao,

This thread will no longer be monitored since we have provided a solution. If you need any additional information from Intel, please submit a new question

Regards,

Syamimi