Synthetic data generation for end-to-end thermal infrared tracking

06/04/2018
by   Lichao Zhang, et al.
0

The usage of both off-the-shelf and end-to-end trained deep networks have significantly improved performance of visual tracking on RGB videos. However, the lack of large labeled datasets hampers the usage of convolutional neural networks for tracking in thermal infrared (TIR) images. Therefore, most state of the art methods on tracking for TIR data are still based on handcrafted features. To address this problem, we propose to use image-to-image translation models. These models allow us to translate the abundantly available labeled RGB data to synthetic TIR data. We explore both the usage of paired and unpaired image translation models for this purpose. These methods provide us with a large labeled dataset of synthetic TIR sequences, on which we can train end-to-end optimal features for tracking. To the best of our knowledge we are the first to train end-to-end features for TIR tracking. We perform extensive experiments on VOT-TIR2017 dataset. We show that a network trained on a large dataset of synthetic TIR data obtains better performance than one trained on the available real TIR data. Combining both data sources leads to further improvement. In addition, when we combine the network with motion features we outperform the state of the art with a relative gain of over 10 showing the efficiency of using synthetic data to train end-to-end TIR trackers.

READ FULL TEXT

page 3

page 5

page 7

page 10

research
08/30/2019

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

We propose an end-to-end tracking framework for fusing the RGB and TIR m...
research
12/04/2017

GANerated Hands for Real-time 3D Hand Tracking from Monocular RGB

We address the highly challenging problem of real-time 3D hand tracking ...
research
10/22/2021

Depth-only Object Tracking

Depth (D) indicates occlusion and is less sensitive to illumination chan...
research
01/11/2021

Multi-Domain Image-to-Image Translation with Adaptive Inference Graph

In this work, we address the problem of multi-domain image-to-image tran...
research
07/22/2023

Self-Supervised and Semi-Supervised Polyp Segmentation using Synthetic Data

Early detection of colorectal polyps is of utmost importance for their t...
research
11/26/2019

Multi-Task Driven Feature Models for Thermal Infrared Tracking

Existing deep Thermal InfraRed (TIR) trackers usually use the feature mo...
research
04/22/2021

Frequency Domain Loss Function for Deep Exposure Correction of Dark Images

We address the problem of exposure correction of dark, blurry and noisy ...

Please sign up or login with your details

Forgot password? Click here to reset