Giter Site home page Giter Site logo

kongbia / lsotb-tir Goto Github PK

View Code? Open in Web Editor NEW

This project forked from qiaoliuhit/lsotb-tir

0.0 0.0 0.0 7 MB

LSOTB-TIR: A Large-Scale High-Diversity Thermal Infrared Object Tracking Benchmark (ACM MM2020)

MATLAB 20.98% M 0.08% Makefile 2.94% HTML 6.95% CSS 1.43% Python 3.36% Clean 0.30% TeX 0.30% Shell 0.08% C 61.18% Objective-C 1.24% Roff 0.49% C++ 0.68%

lsotb-tir's Introduction

LSOTB-TIR: A Large-Scale High-Diversity Thermal Infrared Object Tracking Benchmark

This toolkit is used to evaluate general thermal infrared (TIR) trackers on the TIR object tracking benchmark, LSOTB-TIR, which consists of a large-scale training dataset and an evaluation dataset with a total of 1,400 TIR image sequences and more than 600K frames. To evaluate a TIR tracker on different attributes, we define 4 scenario attributes and 12 challenge attributes in the evaluation dataset. By releasing LSOTB-TIR, we encourage the community to develop deep learning based TIR trackers and evaluate them fairly and comprehensively. Paper, Supplementary materials Alt text

News

  • 2020-08, Our paper is accepted by ACM Multimedia Conference 2020.
  • 2020-11, We update the evaluation dataset because we miss a test sequence 'cat_D_001'.

Characteristics

  • Large-scale: 1400 TIR sequences, 600K+ frames, 730K+ bounding boxes.
  • High-diversity: 12 challenges, 4 scenario, 47 object classes.
  • Contain both training and evaluation data sets.
  • Provide 30+ tracker's evaluation results.

Download dataset and evaluation results

  • Download the dataset and 30+ tracker's evaluation raw results from Dubox using the password: 2fad, if you are not in china.
  • Download the dataset and 30+ tracker's evaluation raw results from Baidu Pan using the password: dr3i, if you are in china.

Usage

  1. Download the evaluation dataset and put it into the sequences folder.
  2. Download the evaluation raw results and put them into the results folder.
  3. Run run_evaluation.m and run_speed.m to draw the result plots.
  4. Configure configTrackers.m and then use main_running_one.m to run your own tracker on the benchmark.

Result's plots

Alt text

Trackers and codes

TIR trackers

  • MMNet. Liu Q, et al. Multi-task driven feature model for thermal infrared tracking, AAAI, 2020. [Github]
  • ECO-stir. Zhang L, et al. Synthetic data generation for end-to-end thermal infrared tracking, TIP, 2019. [Github]
  • MLSSNet. Liu Q, et al, Learning Deep Multi-Level Similarity for Thermal Infrared Object Tracking, TMM, 2020. [Github]
  • HSSNet. Li X, et al, Hierarchical spatial-aware Siamese network for thermal infrared object tracking, KBS, 2019.[Github]
  • MCFTS. Liu Q, et al, Deep convolutional neural networks for thermal infrared object tracking, KBS, 2017. [Github]

RGB trackers

  • ECO. Danelljan M, et al, ECO: efficient convolution operators for tracking, CVPR, 2017. [Github]
  • DeepSTRCF. Li F et al, Learning spatial-temporal regularized correlation filters for visual tracking, CVPR, 2018. [Github]
  • MDNet. Nam H, et al, Learning multi-domain convolutional neural networks for visual tracking, CVPR, 2016. [Github]
  • SRDCF. Danelljan M, et al, Learning spatially regularized correlation filters for visual tracking, ICCV, 2015. [Project]
  • VITAL. Song Y, et al., Vital: Visual tracking via adversarial learning, CVPR, 2018. [Github]
  • TADT. Li X, et al, Target-aware deep tracking, CVPR, 2019. [Github]
  • MCCT. Wang N, et al, Multi-cue correlation filters for robust visual tracking, CVPR, 2018. [Github]
  • Staple. Bertinetto, L, et al, Staple: Complementary learners for real-time tracking, CVPR, 2016. [Github]
  • DSST. Danelljan M, et al, Accurate scale estimation for robust visual tracking, BMVC, 2014. [Github]
  • UDT. Wang N, et al, Unsupervised deep tracking, CVPR, 2019. [Github]
  • CREST. Song Y, et al, Crest: Convolutional residual learning for visual tracking, ICCV, 2017. [Github]
  • SiamFC. Bertinetto, L, et al, Fully-Convolutional Siamese Networks for Object Tracking, ECCVW, 2016. [Github]
  • SiamFC-tri. Dong X, et al, Triplet loss in Siamese network for object tracking, ECCV, 2018. [Github]
  • HDT. Qi Y, et al, Hedged deep tracking, CVPR, 2016. [Project]
  • CFNet. Valmadre, J, et al, End-to-end representation learning for correlation filter based tracking, CVPR, 2017. [Github]
  • HCF. Ma, C, et al, Hierarchical convolutional features for visual tracking, ICCV, 2015. [Github]
  • L1APG. Bao, C, et al, Real time robust L1 tracker using accelerated proximal gradient approach, CVPR, 2012. [Project]
  • SVM. Wang N, et al, Understanding and diagnosing visual tracking systems, ICCV, 2015. [Project]
  • KCF. Henriques, J, et al, High-speed tracking with kernelized correlation filters, TPAMI, 2015. [Project]
  • DSiam. Guo, Q, et al, Learning dynamic siamese network for visual object tracking, ICCV, 2017. [Github]

Contact

Feedbacks and comments are welcome! Feel free to contact us via [email protected] or [email protected]

lsotb-tir's People

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.