• Ubuntu 16.04
• Python 3.6
• Tensorflow 1.10.0

Download the Kitti Object Detection Dataset (image, calib and label) and place it into data/KittiBox. The train-val split train.txt and val.txt are contained in this repository.

Compile the Cython module and download the pretrained model:

python setup.py

Download the .ckpt files and eval results from https://pan.baidu.com/s/1D5exT_8dt-xzWau5cDPTlA (Extracted code：z6a2),and put them in ckpt

The folder should be in the following structure:

data
    KittiBox
        training
            calib
            image_2
            label_2
        train.txt
        val.txt
    model_2D.pkl
    model_3D.data-00000-of-00001
    model_3D.index
    model_3D.meta

Run the evaluating script:

python eval.py

Evaluation results have been put on the "./val_out/val_out",if you want to compute mAp,

./submodules/KittiEvaluation/evaluate_object /path/to/prediction /path/to/gt