LasHeR: A Large-Scale High-Diversity Benchmark for RGBT Tracking (2021-04-27T00:00:00.000000Z)

TL;DR

The unaligned version of LasHeR is released to attract the research interest for alignment-free RGBT tracking, which is a more practical task in real-world applications.

Abstract

RGBT tracking receives a surge of interest in the computer vision community, but this research field lacks a large-scale and high-diversity benchmark dataset, which is essential for both the training of deep RGBT trackers and the comprehensive evaluation of RGBT tracking methods. To this end, we present a <italic>La</italic> rge-<inline-formula> <tex-math notation="LaTeX">${s}$ </tex-math></inline-formula>cale <inline-formula> <tex-math notation="LaTeX">${H}$ </tex-math></inline-formula>igh-diversity <inline-formula> <tex-math notation="LaTeX">$\text{b}{e}$ </tex-math></inline-formula>nchmark for short-term <inline-formula> <tex-math notation="LaTeX">${R}$ </tex-math></inline-formula>GBT tracking (LasHeR) in this work. LasHeR consists of 1224 visible and thermal infrared video pairs with more than 730K frame pairs in total. Each frame pair is spatially aligned and manually annotated with a bounding box, making the dataset well and densely annotated. LasHeR is highly diverse capturing from a broad range of object categories, camera viewpoints, scene complexities and environmental factors across seasons, weathers, day and night. We conduct a comprehensive performance evaluation of 12 RGBT tracking algorithms on the LasHeR dataset and present detailed analysis. In addition, we release the unaligned version of LasHeR to attract the research interest for alignment-free RGBT tracking, which is a more practical task in real-world applications. The datasets and evaluation protocols are available at: <uri>https://github.com/mmic-lcl/Datasets-and-benchmark-code</uri>.

Authors

Chenglong Li

7 papers

Wanli Xue

2 papers

Yaqing Jia

1 papers

TL;DR

Abstract

Authors

References67 items

RGBT Tracking by Trident Fusion Network

RGBT Tracking via Noise-Robust Cross-Modal Ranking

Transformer Tracking

Multimodal Cross-Layer Bilinear Pooling for RGBT Tracking

RGBT Tracking via Multi-Adapter Network with Hierarchical Divergence Loss

Object fusion tracking based on visible and infrared images: A comprehensive review

Challenge-Aware RGBT Tracking

Jointly Modeling Motion and Appearance Cues for Robust RGB-T Tracking

Cross-Modal Pattern-Propagation for RGB-T Tracking

DSiamMFT: An RGB-T fusion tracking method via dynamic Siamese networks using multi-layer feature fusion

Deep Homography Estimation for Dynamic Scenes

Probabilistic Regression for Visual Tracking

Unsupervised Multi-Modal Image Registration via Geometry Preserving Image-to-Image Translation

M5L: Multi-Modal Multi-Margin Metric Learning for RGBT Tracking

Quality-Aware Feature Aggregation Network for Robust RGBT Tracking

Modality-correlation-aware sparse representation for RGB-infrared object tracking

Object Tracking in RGB-T Videos Using Modal-Aware Attention Network and Competitive Learning

SiamFC++: Towards Robust and Accurate Visual Tracking with Target Estimation Guidelines

Deep Adaptive Fusion Network for High Performance RGBT Tracking

The Seventh Visual Object Tracking VOT2019 Challenge Results

Dual-Stream Pyramid Registration Network

Content-Aware Unsupervised Deep Homography Estimation

Multi-Modal Fusion for End-to-End RGB-T Tracking

SiamFT: An RGB-Infrared Fusion Tracking Method via Fully Convolutional Siamese Networks

Dense Feature Aggregation and Pruning for RGBT Tracking

Multi-Adapter RGBT Tracking

Object Fusion Tracking Based on Visible and Infrared Images Using Fully Convolutional Siamese Networks

Thermal infrared and visible sequences fusion tracking based on a hybrid tracking framework with adaptive weighting scheme

Discriminative Fusion Correlation Learning for Visible and Infrared Tracking

Online Non-Negative Multi-Modality Feature Template Learning for RGB-Assisted Infrared Tracking

Learning Discriminative Model Prediction for Tracking

Fast RGB-T Tracking via Cross-Modal Correlation Filters

Learning Modality-Consistency Feature Templates: A Robust RGB-Infrared Tracking System

Weakly Aligned Cross-Modal Learning for Multispectral Pedestrian Detection

Learning Soft-Consistent Correlation Filters for RGB-T Object Tracking

ATOM: Accurate Tracking by Overlap Maximization

CAMEL Dataset for Visual and Thermal Infrared Multiple Object Detection and Tracking

Two-stage modality-graphs regularized manifold ranking for RGB-T tracking

Learning Multi-Domain Convolutional Network for RGB-T Visual Tracking

LaSOT: A High-Quality Benchmark for Large-Scale Single Object Tracking

Cross-Modal Ranking with Soft Consistency and Noisy Labels for Robust RGB-T Tracking

Weakly-supervised convolutional neural networks for multimodal image registration

Synthetic Data Generation for End-to-End Thermal Infrared Tracking

RGB-T Object Tracking: Benchmark and Baseline

Robust Collaborative Discriminative Learning for RGB-Infrared Tracking

TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild

Relative Object Tracking Algorithm Based on Convolutional Neural Network for Visible and Infrared Video Sequences

Fusing two-stream convolutional neural networks for RGB-T object tracking

Weighted Sparse Representation Regularized Graph Learning for RGB-T Object Tracking

Grayscale-Thermal Object Tracking via Multitask Laplacian Sparse Representation

Learning Collaborative Sparse Representation for Grayscale-Thermal Tracking

ViTBAT: Video tracking and behavior annotation tool

Real-Time Grayscale-Thermal Tracking via Laplacian Sparse Representation

Learning Multi-domain Convolutional Neural Networks for Visual Tracking

Comparison of infrared and visible imagery for object tracking: Toward trackers with superior IR performance

Multispectral pedestrian detection: Benchmark dataset and baseline

Fourier Lucas-Kanade Algorithm

An iterative integrated framework for thermal-visible image registration, sensor fusion, and people tracking for video surveillance applications

ORB: An efficient alternative to SIFT or SURF

An improved RANSAC homography algorithm for feature based image mosaic

Background-subtraction using contour-based fusion of thermal and visible imagery

An Iterative Image Registration Technique with an Application to Stereo Vision

Decision-Level Fusion Tracking for Infrared and Visible Spectra Based on Deep Learning

Multi-spectral pedestrian detection based on accumulated object proposal with fully convolutional networks

HEASK: Robust homography estimation based on appearance similarity and keypoint correspondences

ORB: an efﬁcient alternative to SIFT or SURF

pursuing B.Eng. degree in Anhui University, Hefei, China

Field of Study

Journal Information

Name

Page

Volume

Venue Information

Name

Type

URL