How to evaluate (inference) with ssd_300_vgg base trained model

[futurianh1k]

Hi, I followed your instruction, and trained with kitti dataset and ssd_300_vgg base. I could run training, but cannot run evaluation.

I tried evaluation with eval_image_classifier:
python eval_image_classifier.py --alsologtostderr --checkpoint_path=checkpoints\ssd_300_vgg.ckpt --dataset_dir=datasets\kitti --dataset_name=kitti --labels_offset=1 --dataset_split_name=train --model_name=ssd_300_vgg

But, it shows this error: ValueError: Shape must be rank 2 but is rank 0 for 'ssd_preprocessing_train/concat' (op: 'ConcatV2') with input shapes: [1,4], [], [].

Could you please tell me how to solve it? Please show me evaluation or inference command with trained ssd model.

Thank you

[Cuky88]

I couldn't run the eval on custom data, too. But I receive a different error: File "eval_ssd_network.py", line 352, in <module> tf.app.run() File "C:\ProgramData\Miniconda3\envs\tensorflow\lib\site-packages\tensorflow\python\platform\app.py", line 124, in run _sys.exit(main(argv)) File "eval_ssd_network.py", line 276, in main mAP = tf.add_n(list(aps_voc07.values())) / len(aps_voc07) File "C:\ProgramData\Miniconda3\envs\tensorflow\lib\site-packages\tensorflow\python\ops\math_ops.py", line 2120, in add_n raise ValueError("inputs must be a list of at least one Tensor with the " ValueError: inputs must be a list of at least one Tensor with the same dtype and shape

aps_voc07 and aps_voc12 stay empty when I print them out. Could anyone start the eval sofar?