fengyang95 / Pycftrackers
Licence: mit
Python re-implementation of some correlation filter based tracker
Stars: ✭ 313
Programming Languages
python
139335 projects - #7 most used programming language
Labels
Projects that are alternatives of or similar to Pycftrackers
ACViewer
Viewer / Utility for DAT files for the game Asheron's Call
Stars: ✭ 19 (-93.93%)
Mutual labels: dat
dat-shell
Dat shell. Open a remote dat and explore with cd, ls, etc.
Stars: ✭ 62 (-80.19%)
Mutual labels: dat
Blender2NieR
Import/Export WMB/WTP/WTA/DAT/DTT/LAY/COL files for NieR:Automata
Stars: ✭ 115 (-63.26%)
Mutual labels: dat
how-dat-works
Dat documentation [ DEPRECATED - see https://github.com/hypercore-protocol/new-website/tree/master/guides for similar functionality. More info on active projects and modules at https://dat-ecosystem.org/ ]
Stars: ✭ 169 (-46.01%)
Mutual labels: dat
Migrations
SQL database migrations for Golang go-pg and PostgreSQL
Stars: ✭ 251 (-19.81%)
Mutual labels: dat
dat-workshop
How to build web apps using Dat. A workshop by GEUT.
Stars: ✭ 50 (-84.03%)
Mutual labels: dat
decent-dat-intro
a buffet of tiny demos for easing in everyone with developing decentralized apps
Stars: ✭ 67 (-78.59%)
Mutual labels: dat
dat-log
Log History of Dat [ DEPRECATED - More info on active projects and modules at https://dat-ecosystem.org/ ]
Stars: ✭ 12 (-96.17%)
Mutual labels: dat
portal
🔗 zero-config peer-to-peer encrypted live folder syncing that respects your `.gitignore`
Stars: ✭ 284 (-9.27%)
Mutual labels: dat
dat-utils
Utility functions for Beaker Browser's DatArchive API
Stars: ✭ 13 (-95.85%)
Mutual labels: dat
dat
💾 peer-to-peer sharing & live syncronization of files via command line [ DEPRECATED - More info on active projects and modules at https://dat-ecosystem.org/ ]
Stars: ✭ 8,250 (+2535.78%)
Mutual labels: dat
caracara
GEUT LABS. An experimental Dat based collaborative editor.
Stars: ✭ 33 (-89.46%)
Mutual labels: dat
hyperdrive-network-speed
track upload and download speed on a hyperdrive archive [ DEPRECATED - More info on active projects and modules at https://dat-ecosystem.org/ ]
Stars: ✭ 18 (-94.25%)
Mutual labels: dat
Dathttpd
Replaced by Homebase! See https://github.com/beakerbrowser/homebase.
Stars: ✭ 282 (-9.9%)
Mutual labels: dat
hyperhealth
Monitor the health of a hyperfeed (hypercore or hyperdrive), including peer count and peer mirror percentages.
Stars: ✭ 38 (-87.86%)
Mutual labels: dat
paperslip
share hard-to-transmit snippets with easy-to-pronounce names using dht magic
Stars: ✭ 37 (-88.18%)
Mutual labels: dat
pyCFTrackers
Python re-implementation of some correlation filter based tracker, and all of these algorithms are implemented based on the official Matlab code. All the code has been tested on Ubuntu 16.04, Python 3.5. I use pysot-toolkit to eval the performance on OTB and VOT.
- [x] MOSSE
- [x] CSK
- [x] CN
- [x] KCF/DCF
- [x] DSST/DSST-LP
- [x] SAMF
- [x] Staple
- [x] Staple-CA
- [x] BACF
- [x] CSR-DCF/CSR-DCF-LP
- [x] DAT This tracker is not based on CF, I re-implement it just for understanding Staple better
- [x] LDES
- [x] MKCFup/MKCFup-LP
- [x] MCCTH
- [x] STRCF
Trackers from other repos - [x] ECO/ECO-HC from pyECO
- [x] OPENCV-CSRDCF(Note that trackers implemented in opencv get lower score because they will return a success flag and if this flag is False, the bbox will be set to (0,0,0,0))
Install
git clone https://github.com/wwdguu/pyCFTrackers.git && cd pyCFTrackers
export pyCFTrackers=$PWD
pip install -r requirements.txt
cd lib/eco/features/
python setup.py build_ext --inplace
cd ../../..
cd lib/pysot/utils/
python setup.py build_ext --inplace
cd ../../..
export PYTHONPATH=$PWD:$PYTHONPATH
Get Dataset
You can follow the instructions in the following repo to get VOT2016,VOT2018 and OTB100 dataset.
trackdat
Then get the json files according to pysot-toolkit
Then put the data in the dataset dir.
Demo
cd examples
python cf_demo.py
Eval on OTB and VOT
cd eval
python get_vot2016_result.py
python get_vot2018_result.py
python ope_otb.py
python eval_VOT2016.py
python eval_VOT2018.py
python eval_OTB.py
OTB result
OTB-100
OTB100 Success Plot | OTB100 Precision Plot |
---|---|
OTB-2013
OTB2013 Success Plot | OTB2013 Precision Plot |
---|---|
VOT result
VOT2018
---------------------------------------------------------------
| Tracker Name | Accuracy | Robustness | Lost Number | EAO |
---------------------------------------------------------------
| ECO | 0.485 | 0.403 | 86.0 | 0.224 |
| CSRDCF | 0.492 | 0.501 | 107.0 | 0.210 |
| ECO-HC | 0.500 | 0.473 | 101.0 | 0.207 |
| CSRDCF-LP | 0.503 | 0.553 | 118.0 | 0.199 |
| Staple | 0.524 | 0.665 | 142.0 | 0.179 |
| LDES | 0.528 | 0.684 | 146.0 | 0.175 |
| MCCTH-Staple | 0.535 | 0.684 | 146.0 | 0.172 |
| OPENCV-CSRDCF | 0.486 | 0.651 | 139.0 | 0.170 |
| BACF | 0.511 | 0.674 | 144.0 | 0.169 |
| DAT | 0.477 | 0.777 | 166.0 | 0.158 |
| STRCF | 0.483 | 0.679 | 145.0 | 0.152 |
| CN | 0.439 | 1.100 | 235.0 | 0.112 |
| SAMF | 0.499 | 1.147 | 245.0 | 0.110 |
| DSST | 0.492 | 1.222 | 261.0 | 0.107 |
| DSST-LP | 0.512 | 1.260 | 269.0 | 0.103 |
| DCF | 0.463 | 1.246 | 266.0 | 0.099 |
| KCF | 0.463 | 1.330 | 284.0 | 0.094 |
| CSK | 0.418 | 1.386 | 296.0 | 0.090 |
| MOSSE | 0.378 | 1.967 | 420.0 | 0.063 |
---------------------------------------------------------------
VOT2016
---------------------------------------------------------------
| Tracker Name | Accuracy | Robustness | Lost Number | EAO |
---------------------------------------------------------------
| ECO | 0.564 | 0.256 | 55.0 | 0.336 |
| MCCTH-Staple | 0.574 | 0.359 | 77.0 | 0.303 |
| Staple | 0.560 | 0.387 | 83.0 | 0.299 |
| ECO-HC | 0.532 | 0.350 | 75.0 | 0.293 |
| CSRDCF | 0.542 | 0.359 | 77.0 | 0.273 |
| LDES | 0.577 | 0.419 | 90.0 | 0.272 |
| CSRDCF-LP | 0.548 | 0.354 | 76.0 | 0.272 |
| BACF | 0.521 | 0.405 | 87.0 | 0.252 |
| OPENCV-CSRDCF | 0.521 | 0.438 | 94.0 | 0.239 |
| STRCF | 0.520 | 0.415 | 89.0 | 0.239 |
| DAT | 0.474 | 0.503 | 108.0 | 0.232 |
| SAMF | 0.544 | 0.639 | 137.0 | 0.193 |
| DSST-LP | 0.543 | 0.727 | 156.0 | 0.180 |
| CN | 0.468 | 0.653 | 140.0 | 0.178 |
| DSST | 0.531 | 0.732 | 157.0 | 0.177 |
| DCF | 0.474 | 0.704 | 151.0 | 0.171 |
| KCF | 0.469 | 0.718 | 154.0 | 0.171 |
| CSK | 0.433 | 0.886 | 190.0 | 0.139 |
| MOSSE | 0.388 | 1.244 | 267.0 | 0.096 |
---------------------------------------------------------------
License.
Licensed under an MIT license.
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].