Skip to content

Bref: SNS-CF is developed by Ntwari Thierry during a master course at Chungang University, South Korea under IPIS Lab (Information Processing and Intelligent Systems Laboratory). It implements several visual tracking algorithms including a brand new SNS-CF. Enjoy the ride...

License

Notifications You must be signed in to change notification settings

jonthierry/SNS-CF

Repository files navigation

SNS-CF: Siamese Network and Spatially Semantic Features for Object Tracking.

Bref: SNS-CF is developed by Ntwari Thierry during a master course at Chungang University, South Korea under IPIS Lab (Information Processing and Intelligent Systems Laboratory). It implements several visual tracking algorithms including a brand new SNS-CF. Enjoy the ride...

MDPI sensors published manuscript: https://doi.org/10.3390/s20174881

Installation

Create the conda environment

conda env create -f environment.yml
conda activate SNS-CF

Build the vot toolkits

bash benchmark/make_toolkits.sh

Get the Models

The .pth for SiamFC is already in the repo since it is small. The SiamRPN and SiamMask models need to be downloaded and moved to their respective sub-folder.

SiamRPN Get model here → move to ./trackers/SiamRPN/model.pth

SiamMask

Download the model and move to sub-folder

wget http://www.robots.ox.ac.uk/~qwang/SiamMask_VOT.pth
mv SiamMask_VOT.pth trackers/SiamMask/model.pth

Get the Datasets

Run the download script to get the OTB2015 and VOT2018 datasets.

cd data/
bash get_test_data.sh

Reproduce the benchmark results

To reproduce the benchmark results from the paper run for example:

On VOT2018, with SNS-CF

python bench.py -d VOT2018 -t SiamRPN

On VOT2018, with SNS-CF and visualizations on

python bench.py -d VOT2018 -t SiamRPN --viz

On VOT2018, with SNS-CF and verbose mode on

python bench.py -d VOT2018 -t SiamRPN --verbose

On VOT2018, with SNS-CF and with a specific video

python bench.py -d VOT2018 -t SiamRPN --spec_video VIDEO_NAME
Option     | |
-------- | ---
``-d DATASET, --dataset DATASET``| ``(Required) Dataset on which the benchmark is run [VOT2018, OTB2015]``
``-t TRACKER, --tracker TRACKER``| ``(Required) Name of the tracker [SiamFC, SiamRPN, SiamMask]``
``-v, --viz``| `` Show the tracked scene, the stored templated and the modulated view``
``--verbose``| `` Print additional info about THOR``
``--spec_video VIDEO_NAME``| `` Pick a specific video by name, e.g. "lemming" on OTB2015``

Webcam demo

To run **SNS-CF** on your webcam / camera, run the following commands #### With SNS-CF (ensemble) enabled, using SiamRPN ```shell script python webcam_demo.py --tracker SiamRPN ```

With SNS-CF and visualizations on

python webcam_demo.py --tracker SiamRPN --viz

With SNS-CF and verbose mode on

python webcam_demo.py --tracker SiamRPN --verbose
Option     | |
-------- | ---
``-t TRACKER, --tracker TRACKER``| ``(Required) Name of the tracker [SiamFC, SiamRPN, SiamMask]``
``-v, --viz``| `` Show the stored templated and the modulated view``
``--verbose``| `` Print additional info about THOR``

Acknowledgments

The code (with some modifications) and models of the trackers are from the following linked repositories:

SiamFC, SiamRPN, SiamMask, THOR, SNS-CF


******************************** ©jonthierry 2020 ***********************************

About

Bref: SNS-CF is developed by Ntwari Thierry during a master course at Chungang University, South Korea under IPIS Lab (Information Processing and Intelligent Systems Laboratory). It implements several visual tracking algorithms including a brand new SNS-CF. Enjoy the ride...

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published