Python Implementation of ECO
Run demo
cd pyECO/eco/features/
python setup.py build_ext --inplace
pip install numpy scipy python-opencv glob pandas pillow
# if you want to use deep feature
pip install mxnet-cu80(or 90 according to your cuda version)
pip install cupy-cu80(or 90 according to your cuda version)
cd pyECO/
python bin/demo_ECO_hc.py --video_dir path/to/videoConvert to deep feature version
uncomment eco/config/config.py at line5 and comment eco/config/config.py at line 6
Benchmark results
OTB100
| Tracker | AUC | Speed |
|---|---|---|
| ECO_deep | 68.7(vs 69.1) | 6~8fps on GTX 1080 Ti |
| ECO_hc | 65.2(vs 65.0) | 40~60fps on Intel i5 |
Visualization Results
Note
we use ResNet50 feature instead of the original imagenet-vgg-m-2048
code tested on mac os 10.13 and python 3.6, ubuntu 16.04 and python 3.6
Reference
[1] Danelljan, Martin and Bhat, Goutam and Shahbaz Khan, Fahad and Felsberg, Michael ​ ECO: Efficient Convolution Operators for Tracking ​ In Conference on Computer Vision and Pattern Recognition (CVPR), 2017