train own dataset, but Found 0 boxes for img

[yfq]

I use 40 classes datasets to train new model by 'train_bottleneck.py' (or train.py), epochs=5, end loss=30, But use new model predict mp4 video, shows Found 0 boxes for img all the time.
Can you help me?

[huichen98]

The problem is your voc_annotation.py file. In this file，you should modify your classes which belong to you！