I am training an object detection model out of the TensorFlow Model Garden (namely, 'http://download.tensorflow.org/models/object_detection/tf2/20200711/ssd_mobilenet_v2_fpnlite_320x320_coco17_tpu-8.tar.gz'). I configure my data augmentation in the pipline.config file. The full list of data augmentation options are in the preprocessor.proto under the tensorflow/models/research/object_detection/protos.
Here is my pipeline.config content:
model {
ssd {
num_classes: 2
image_resizer {
fixed_shape_resizer {
height: 320
width: 320
}
}
feature_extractor {
type: "ssd_mobilenet_v2_fpn_keras"
depth_multiplier: 1.0
min_depth: 16
conv_hyperparams {
regularizer {
l2_regularizer {
weight: 4e-05
}
}
initializer {
random_normal_initializer {
mean: 0.0
stddev: 0.01
}
}
activation: RELU_6
batch_norm {
decay: 0.997
scale: true
epsilon: 0.001
}
}
use_depthwise: true
override_base_feature_extractor_hyperparams: true
fpn {
min_level: 3
max_level: 7
additional_layer_depth: 128
}
}
box_coder {
faster_rcnn_box_coder {
y_scale: 10.0
x_scale: 10.0
height_scale: 5.0
width_scale: 5.0
}
}
matcher {
argmax_matcher {
matched_threshold: 0.5
unmatched_threshold: 0.5
ignore_thresholds: false
negatives_lower_than_unmatched: true
force_match_for_each_row: true
use_matmul_gather: true
}
}
similarity_calculator {
iou_similarity {
}
}
box_predictor {
weight_shared_convolutional_box_predictor {
conv_hyperparams {
regularizer {
l2_regularizer {
weight: 4e-05
}
}
initializer {
random_normal_initializer {
mean: 0.0
stddev: 0.01
}
}
activation: RELU_6
batch_norm {
decay: 0.997
scale: true
epsilon: 0.001
}
}
depth: 128
num_layers_before_predictor: 4
kernel_size: 3
class_prediction_bias_init: -4.6
share_prediction_tower: true
use_depthwise: true
}
}
anchor_generator {
multiscale_anchor_generator {
min_level: 3
max_level: 7
anchor_scale: 4.0
aspect_ratios: 1.0
aspect_ratios: 2.0
aspect_ratios: 0.5
scales_per_octave: 2
}
}
post_processing {
batch_non_max_suppression {
score_threshold: 1e-08
iou_threshold: 0.6
max_detections_per_class: 100
max_total_detections: 100
use_static_shapes: false
}
score_converter: SIGMOID
}
normalize_loss_by_num_matches: true
loss {
localization_loss {
weighted_smooth_l1 {
}
}
classification_loss {
weighted_sigmoid_focal {
gamma: 2.0
alpha: 0.25
}
}
classification_weight: 1.0
localization_weight: 1.0
}
encode_background_as_zeros: true
normalize_loc_loss_by_codesize: true
inplace_batchnorm_update: true
freeze_batchnorm: false
}
}
train_config {
batch_size: 4
random_horizontal_flip {
keypoint_flip_permutation: 1
keypoint_flip_permutation: 0
keypoint_flip_permutation: 2
keypoint_flip_permutation: 3
keypoint_flip_permutation: 5
keypoint_flip_permutation: 4
probability: 0.5
}
}
data_augmentation_options {
random_vertical_flip {
keypoint_flip_permutation: 1
keypoint_flip_permutation: 0
keypoint_flip_permutation: 2
keypoint_flip_permutation: 3
keypoint_flip_permutation: 5
keypoint_flip_permutation: 4
probability: 0.5
}
}
data_augmentation_options {
normalize_image{
original_minval: 0.0
original_maxval: 255.0
target_minval: -1.0
target_maxval: 1.0
}
}
data_augmentation_options {
random_image_scale{
min_scale_ratio: 0.8
max_scale_ratio: 2.2
}
}
sync_replicas: true
optimizer {
momentum_optimizer {
learning_rate {
cosine_decay_learning_rate {
learning_rate_base: 0.08
total_steps: 50000
warmup_learning_rate: 0.026666
warmup_steps: 1000
}
}
momentum_optimizer_value: 0.9
}
use_moving_average: false
}
fine_tune_checkpoint: "C:/Users/abukwah/source/repos/Edge_Detection/Tensorflow/workspace/pre-trained-models/ssd_mobilenet_v2_fpnlite_320x320_coco17_tpu-8/checkpoint/ckpt-0"
num_steps: 50000
startup_delay_steps: 0.0
replicas_to_aggregate: 8
max_number_of_boxes: 100
unpad_groundtruth_tensors: false
fine_tune_checkpoint_type: "detection"
fine_tune_checkpoint_version: V2
}
train_input_reader {
label_map_path: "C:/Users/Katzoos/source/repos/Edge_Detection/Tensorflow/workspace/annotations/label_map.pbtxt"
tf_record_input_reader {
input_path: "C:/Users/Katzoos/source/repos/Edge_Detection/Tensorflow/workspace/annotations/train.record"
}
}
eval_config {
metrics_set: "coco_detection_metrics"
use_moving_averages: false
}
eval_input_reader {
label_map_path: "C:/Users/Katzoos/source/repos/Edge_Detection/Tensorflow/workspace/annotations/label_map.pbtxt"
shuffle: false
num_epochs: 1
tf_record_input_reader {
input_path: "C:/Users/Katzoos/source/repos/Edge_Detection/Tensorflow/workspace/annotations/test.record"
}
}
My question is how can I visualize the results of the data augmentation that is part of the training pipeline? And can I see it after the training is done (is saved somewhere) or should I edit the code in the Tensorflow\models\research\object_detection\model_main_tf2.py file?
I am using Tensorflow 2.10.0, Python 3.9.15, CUDA 11.2, CuDNN 8.1.
I tried to find answers on the Tensorflow Model Garden documentation and couldn't find answer to my question.
