Donate. I desperately need donations to survive due to my health

Get paid by answering surveys Click here

Click here to donate

Remote/Work from Home jobs

Can not train my dataset in Tensorflow [Unable to decode bytes as JPEG, PNG, GIF, or BMP]

on

Hello I'am new on Stackoverflow.
I have a problem training my own dataset with the ssd_mobilenet_v2_coco Model.

I have followed all the steps on this site: Link
And nothing gets wrong.

My images are labeled with LabelImg.

generate_tfrecord.py: Link to code

My label_map.pbtxt:

item {
  id: 1
  name: 'person'
  display_name: 'Person'
}

My config file:Link to config file

My images have a max size of 640x480 or 480x640.

I have tested the filetype of all images, all have the filetype JPEG. Link to code

every time I execute this command python model_main.py --logtostderr --train_dir=/home/bartek/Downloads/person/data/ --pipeline_config_path=training/pipeline.config in the directory models/research/object_detection I get this error:

tensorflow.python.framework.errors_impl.InvalidArgumentError: assertion failed: [Unable to decode bytes as JPEG, PNG, GIF, or BMP]
[[{{node case/cond/cond_jpeg/decode_image/cond_jpeg/cond_png/cond_gif/Assert_1/Assert}} = Assert[T=[DT_STRING], summarize=3, _device="/device:CPU:0"](case/cond/cond_jpeg/decode_image/cond_jpeg/cond_png/cond_gif/is_bmp, case/cond/cond_jpeg/decode_image/cond_jpeg/cond_png/cond_gif/Assert_1/Assert/data_0)]] [[node IteratorGetNext (defined at model_main.py:105) = IteratorGetNextoutput_shapes=[[12], [12,300,300,3], [12,2], [12,3], [12,100], [12,100,4], [12,100,1], [12,100,1], [12,100], [12,100], [12,100], [12]], output_types=[DT_INT32, DT_FLOAT, DT_INT32, DT_INT32, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_INT32, DT_BOOL, DT_FLOAT, DT_INT32], _device="/job:localhost/replica:0/task:0/device:CPU:0"]]"

Full error trace

Now I do not know what the problem is!?

I hope someone can help me!

Thanks

Comments

Post a Comment