Tensorflow SSD - mobilenet TypeError: 'NoneType' object cannot be interpreted as an integer

[ingenious-hans]

This error is prompting when I run

python generate_tfrecord.py --csv_input=data/train_labels.csv --output_path=data/train.record I am trying to build a custom object detector using Single Shot Detector mobilenet

The error when building TF records generate_tfrecord.py

python generate_tfrecord.py --csv_input=data/train_labels.csv --output_path=data/train.record

The CSV file has columns: filename, width, eight, class, xmin, ymin, xmax, ymax and the image files are well placed

[zzanyoung]

In my case, I encountered a similar error and ended up using roboflow to generate labels and tfrecords