I am creating tree detector using Tensoflow Object Detection API with the help of following tutorial : https://www.youtube.com/watch?v=a1br6gW-8Ss
My problem is that the model is overfitting the training set. How can I avoid it ? My training set is composed of 130 images and on average 4 tree is labeled in image. There are 7 species at all. The model I use is : 'Faster R-CNN ResNet50 V1 640x640' pre-trained on COCO dataset..(Link is below) https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/tf2_detection_zoo.md In tutorial tutor uses the following model 'EfficientDet D0 512x512' but that model also overfits my dataset so I tried to use 'Faster R-CNN ResNet50 V1 640x640'.
The techniques I use to reduce overfitting is
- dropout(rate=0.675)
- data augmentation(random_crop)
- Low learning rate(learning_rate_base: .008 , total_steps: 25000 .warmup_learning_rate: .0001)
- original images dimensions were 860x860 and I resized them tp 640x640 and changed bounding box coordinates also in xml.
What should I add or change in model configuration to avoid overfitting ?
Here is my model configuration file:
# Faster R-CNN with Resnet-50 (v1) with 640x640 input resolution
# Trained on COCO, initialized from Imagenet classification checkpoint
#
# Train on TPU-8
#
# Achieves 29.3 mAP on COCO17 Val
model {
faster_rcnn {
num_classes: 7
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 {
scales: [0.25, 0.5, 1.0, 2.0]
aspect_ratios: [0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5]
height_stride: 8
width_stride: 8
}
}
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 {
use_dropout: true
dropout_keep_probability: 0.675
fc_hyperparams {
op: FC
regularizer {
l2_regularizer {
weight: 0.0
}
}
initializer {
variance_scaling_initializer {
factor: 1.0
uniform: true
mode: FAN_AVG
}
}
}
share_box_across_classes: true
}
}
second_stage_post_processing {
batch_non_max_suppression {
score_threshold: 0.2
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_static_shapes: true
use_matmul_crop_and_resize: true
clip_anchors_to_image: true
use_static_balanced_label_sampler: true
use_matmul_gather_in_matcher: true
}
}
train_config: {
batch_size: 4
sync_replicas: true
startup_delay_steps: 0
replicas_to_aggregate: 8
num_steps: 25000
optimizer {
momentum_optimizer: {
learning_rate: {
cosine_decay_learning_rate {
learning_rate_base: .008
total_steps: 25000
warmup_learning_rate: .0001
warmup_steps: 2000
}
}
momentum_optimizer_value: 0.9
}
use_moving_average: false
}
fine_tune_checkpoint_version: V2
fine_tune_checkpoint: "faster_rcnn_resnet50_v1_640x640_coco17_tpu-8/checkpoint/ckpt-0"
fine_tune_checkpoint_type: "detection"
data_augmentation_options {
random_crop_image {
min_object_covered: 0.0
min_aspect_ratio: 0.75
max_aspect_ratio: 3.0
min_area: 0.75
max_area: 1.0
overlap_thresh: 0.0
}
}
max_number_of_boxes: 100
unpad_groundtruth_tensors: false
use_bfloat16: true # works only on TPUs
}
train_input_reader: {
label_map_path: "images/labelmap.pbtxt"
tf_record_input_reader {
input_path: "train.record"
}
}
eval_config: {
metrics_set: "coco_detection_metrics"
use_moving_averages: false
batch_size: 1;
}
eval_input_reader: {
label_map_path: "images/labelmap.pbtxt"
shuffle: false
num_epochs: 1
tf_record_input_reader {
input_path: "test.record"
}
}