How to visualize a TFRecord?

Question

I was asked this on another forum but thought I'd post it here for anyone that is having trouble with TFRecords.

TensorFlow's Object Detection API can produce strange behavior if the labels in the TFRecord file do not align with the labels in your labels.pbtxt file. It will run, loss will likely decrease but the network will not produce good detections.

Also, I for one always get confused between X-Y, row-col space, and so I always like to double check to make sure that my annotations are actually annotating the right parts of the image.

The best way I've found to do this is by decoding the TFRecord and plotting it with TF tools. Here's some code below:

import matplotlib.pyplot as plt
import numpy as np
import tensorflow as tf
from object_detection.utils import visualization_utils as vu
from object_detection.protos import string_int_label_map_pb2 as pb
from object_detection.data_decoders.tf_example_decoder import TfExampleDecoder as TfDecoder
from google.protobuf import text_format
def main(tfrecords_filename, label_map=None):
    if label_map is not None:
        label_map_proto = pb.StringIntLabelMap()
        with tf.gfile.GFile(label_map,'r') as f:
            text_format.Merge(f.read(), label_map_proto)
            class_dict = {}
            for entry in label_map_proto.item:
                class_dict[entry.id] = {'name':entry.display_name}
    sess = tf.Session()
    decoder = TfDecoder(label_map_proto_file=label_map, use_display_name=False)
    sess.run(tf.tables_initializer())
    for record in tf.python_io.tf_record_iterator(tfrecords_filename):
        example = decoder.decode(record)
        host_example = sess.run(example)
        scores = np.ones(host_example['groundtruth_boxes'].shape[0])
        vu.visualize_boxes_and_labels_on_image_array( 
            host_example['image'],                                               
            host_example['groundtruth_boxes'],                                                     
            host_example['groundtruth_classes'],
            scores,
            class_dict,
            max_boxes_to_draw=None,
            use_normalized_coordinates=True)
plt.imshow(host_example['image'])
plt.show()

Answer 1

If you'd like to check bounding boxes/labels visually, you can check this TFRecord Viewer: https://github.com/sulc/tfrecord-viewer