Home

Awesome

SfMLearner Pytorch version

This codebase implements the system described in the paper:

Unsupervised Learning of Depth and Ego-Motion from Video

Tinghui Zhou, Matthew Brown, Noah Snavely, David G. Lowe

In CVPR 2017 (Oral).

See the project webpage for more details.

Original Author : Tinghui Zhou (tinghuiz@berkeley.edu) Pytorch implementation : Clément Pinard (clement.pinard@ensta-paristech.fr)

sample_results

Preamble

This codebase was developed and tested with Pytorch 1.0.1, CUDA 10 and Ubuntu 16.04. Original code was developped in tensorflow, you can access it here

Prerequisite

pip3 install -r requirements.txt

or install manually the following packages :

pytorch >= 1.0.1
pebble
matplotlib
imageio
scipy
scikit-image
argparse
tensorboardX
blessings
progressbar2
path.py

Note

Because it uses latests pytorch features, it is not compatible with anterior versions of pytorch.

If you don't have an up to date pytorch, the tags can help you checkout the right commits corresponding to your pytorch version.

What has been done

Differences with official Implementation

Preparing training data

Preparation is roughly the same command as in the original code.

For KITTI, first download the dataset using this script provided on the official website, and then run the following command. The --with-depth option will save resized copies of groundtruth to help you setting hyper parameters. The --with-pose will dump the sequence pose in the same format as Odometry dataset (see pose evaluation)

python3 data/prepare_train_data.py /path/to/raw/kitti/dataset/ --dataset-format 'kitti_raw' --dump-root /path/to/resulting/formatted/data/ --width 416 --height 128 --num-threads 4 [--static-frames /path/to/static_frames.txt] [--with-depth] [--with-pose]

For Cityscapes, download the following packages: 1) leftImg8bit_sequence_trainvaltest.zip, 2) camera_trainvaltest.zip. You will probably need to contact the administrators to be able to get it. Then run the following command

python3 data/prepare_train_data.py /path/to/cityscapes/dataset/ --dataset-format 'cityscapes' --dump-root /path/to/resulting/formatted/data/ --width 416 --height 171 --num-threads 4

Notice that for Cityscapes the img_height is set to 171 because we crop out the bottom part of the image that contains the car logo, and the resulting image will have height 128.

Training

Once the data are formatted following the above instructions, you should be able to train the model by running the following command

python3 train.py /path/to/the/formatted/data/ -b4 -m0.2 -s0.1 --epoch-size 3000 --sequence-length 3 --log-output [--with-gt]

You can then start a tensorboard session in this folder by

tensorboard --logdir=checkpoints/

and visualize the training progress by opening https://localhost:6006 on your browser. If everything is set up properly, you should start seeing reasonable depth prediction after ~30K iterations when training on KITTI.

Evaluation

Disparity map generation can be done with run_inference.py

python3 run_inference.py --pretrained /path/to/dispnet --dataset-dir /path/pictures/dir --output-dir /path/to/output/dir

Will run inference on all pictures inside dataset-dir and save a jpg of disparity (or depth) to output-dir for each one see script help (-h) for more options.

Disparity evaluation is avalaible

python3 test_disp.py --pretrained-dispnet /path/to/dispnet --pretrained-posenet /path/to/posenet --dataset-dir /path/to/KITTI_raw --dataset-list /path/to/test_files_list

Test file list is available in kitti eval folder. To get fair comparison with Original paper evaluation code, don't specify a posenet. However, if you do, it will be used to solve the scale factor ambiguity, the only ground truth used to get it will be vehicle speed which is far more acceptable for real conditions quality measurement, but you will obviously get worse results.

Pose evaluation is also available on Odometry dataset. Be sure to download both color images and pose !

python3 test_pose.py /path/to/posenet --dataset-dir /path/to/KITIT_odometry --sequences [09]

ATE (Absolute Trajectory Error) is computed as long as RE for rotation (Rotation Error). RE between R1 and R2 is defined as the angle of R1*R2^-1 when converted to axis/angle. It corresponds to RE = arccos( (trace(R1 @ R2^-1) - 1) / 2). While ATE is often said to be enough to trajectory estimation, RE seems important here as sequences are only seq_length frames long.

Pretrained Nets

Avalaible here

Arguments used :

python3 train.py /path/to/the/formatted/data/ -b4 -m0 -s2.0 --epoch-size 1000 --sequence-length 5 --log-output --with-gt

Depth Results

Abs RelSq RelRMSERMSE(log)Acc.1Acc.2Acc.3
0.1811.3416.2360.2620.7330.9010.964

Pose Results

5-frames snippets used

Seq. 09Seq. 10
ATE0.0179 (std. 0.0110)0.0141 (std. 0.0115)
RE0.0018 (std. 0.0009)0.0018 (std. 0.0011)

Discussion

Here I try to link the issues that I think raised interesting questions about scale factor, pose inference, and training hyperparameters

Other Implementations

TensorFlow by tinghuiz (original code, and paper author)