Using synthetic data for person tracking under adverse weather conditions

Kerim, Abdulrahman; Çelikcan, Ufuk; Erdem, Erkut; Erdem, Aykut

doi:10.1016/j.imavis.2021.104187

Cited by 15 publications

(7 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A further and well-known issue could arise from variations in weather conditions and daytime lighting, affecting image illumination and colours. Nevertheless, synthetic images can be an effective solution to mitigate this issue: for instance, synthetic images simulating lighting and colour variations and specific weather conditions can be generated, and different models can be trained for specific conditions, which can then be easily selected by end-users depending on the particular environmental conditions [30]. Another interesting issue for future investigations is to improve the realism of synthetic images using computer graphics tools or GANs [11], to transfer the style of the target cameras to pedestrian images in the gallery.…”

Section: Discussionmentioning

confidence: 99%

Scene-specific crowd counting using synthetic training images

Delussu

Putzu²,

Fumera³

2022

Pattern Recognition

View full text Add to dashboard Cite

Section: Discussionmentioning

confidence: 99%

Scene-specific crowd counting using synthetic training images

Delussu

Putzu²,

Fumera³

2022

Pattern Recognition

View full text Add to dashboard Cite

“…The detection and tracking of occluded body joints along with a new dataset obtained from the Grand Theft Auto V (GTA V) videogame are presented in [23]. Taking advantage of the same simulator, [7] makes available a benchmark for multi-target tracking in a multi-camera system with and without overlapping cameras, while [8] with NOVA focuses on evaluating the robustness of tracking methods under adverse weather conditions. Other works, such as [24], [25], tap into the effortlessness of getting labeled data with simulators and propose domain adaptation algorithms from synthetic to real-world data for action recognition and re-identification, respectively.…”

Section: Related Workmentioning

confidence: 99%

A Self-Adaptive Gallery Construction Method for Open-World Person Re-Identification

Casao

Azagra

Murillo

et al. 2023

Sensors

View full text Add to dashboard Cite

Person re-identification, or simply re-id, is the task of identifying again a person who has been seen in the past by a perception system. Multiple robotic applications, such as tracking or navigate-and-seek, use re-identification systems to perform their tasks. To solve the re-id problem, a common practice consists in using a gallery with relevant information about the people already observed. The construction of this gallery is a costly process, typically performed offline and only once because of the problems associated with labeling and storing new data as they arrive in the system. The resulting galleries from this process are static and do not acquire new knowledge from the scene, which is a limitation of the current re-id systems to work for open-world applications. Different from previous work, we overcome this limitation by presenting an unsupervised approach to automatically identify new people and incrementally build a gallery for open-world re-id that adapts prior knowledge with new information on a continuous basis. Our approach performs a comparison between the current person models and new unlabeled data to dynamically expand the gallery with new identities. We process the incoming information to maintain a small representative model of each person by exploiting concepts of information theory. The uncertainty and diversity of the new samples are analyzed to define which ones should be incorporated into the gallery. Experimental evaluation in challenging benchmarks includes an ablation study of the proposed framework, the assessment of different data selection algorithms that demonstrate the benefits of our approach, and a comparative analysis of the obtained results with other unsupervised and semi-supervised re-id methods.

show abstract

“…No quantitative experiments on real-world datasets were performed. Kerim et al [55], on the other hand, created their own rendering engine (NOVA) based on Unity to allow researchers with no experience in computer graphics to generate high quality datasets with accurate and dense annotations. The authors collected a real-world dataset named PTAW172Real, and they used NOVA to generate a synthetic one called PTAW217Synt.…”

Section: Simulationmentioning

confidence: 99%

Survey on Videos Data Augmentation for Deep Learning Models

Cauli

Recupero

2022

Future Internet

View full text Add to dashboard Cite

In most Computer Vision applications, Deep Learning models achieve state-of-the-art performances. One drawback of Deep Learning is the large amount of data needed to train the models. Unfortunately, in many applications, data are difficult or expensive to collect. Data augmentation can alleviate the problem, generating new data from a smaller initial dataset. Geometric and color space image augmentation methods can increase accuracy of Deep Learning models but are often not enough. More advanced solutions are Domain Randomization methods or the use of simulation to artificially generate the missing data. Data augmentation algorithms are usually specifically designed for single images. Most recently, Deep Learning models have been applied to the analysis of video sequences. The aim of this paper is to perform an exhaustive study of the novel techniques of video data augmentation for Deep Learning models and to point out the future directions of the research on this topic.

show abstract

Using synthetic data for person tracking under adverse weather conditions

Cited by 15 publications

References 35 publications

Scene-specific crowd counting using synthetic training images

Scene-specific crowd counting using synthetic training images

A Self-Adaptive Gallery Construction Method for Open-World Person Re-Identification

Survey on Videos Data Augmentation for Deep Learning Models

Contact Info

Product

Resources

About