Automatic Adaptation of Object Detectors to New Domains Using Self-Training

To proactively assess the losses caused by the deterioration of metro tunnels during the operational period, a new method, the cloud model‐based random forests (CRFs), is proposed to discuss the inconsistencies induced by mapping the monitoring data into the health rate of the metro tunnels. On top of the CRF, a self‐training framework is introduced to improve the predictive accuracy and the stability of the CRF by adding more unlabeled data. The main contribution of this paper is proposing a novel CRF model along with a semisupervised approach to overcome the inapplicability of the random forests algorithm in an inconsistent small database, and to reduce the substantial time and cost for expert annotation. The results indicate that (1) the proposed CRF achieves higher accuracy and stability than random forests in predictions; (2) the CRF outperforms the other state‐of‐the‐art methods even in a small database; (3) the self‐training improved CRF keeps highly precise when the ratio of labeled to unlabeled data is no less than 1:11.4. In this study, the suggested ratio of labeled and unlabeled data is no lower than 1:5.7 to reduce the risk of wrongly forecasting a seriously damaged tunnel section as a slightly damaged tunnel section.

show abstract

“…RoyChowdhury et al. (2019) trained a self‐training model based on web videos to detect pedestrians’ faces under various scenarios. Rogers et al.…”

Section: Methodsmentioning

confidence: 99%

Tunnel condition assessment via cloud model‐based random forests and self‐training approach

Zhu

Guo

et al. 2020

Computer aided Civil Eng

View full text Add to dashboard Cite

show abstract

“…Domain Adaptation for Object Detection. Unsupervised domain adaptation for object detection has recently gained interest [17], [46], [20], [47], [18], [48], [49], [50], [22], [21]. Faster R-CNN has been adapted for domain adaptation by aligning the distributions of the last convolutional feature map and the region features [17].…”

Section: Related Workmentioning

confidence: 99%

AFAN: Augmented Feature Alignment Network for Cross-Domain Object Detection

Wang

Liao

Shao

2021

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

Unsupervised domain adaptation for object detection is a challenging problem with many real-world applications. Unfortunately, it has received much less attention than supervised object detection. Models that try to address this task tend to suffer from a shortage of annotated training samples. Moreover, existing methods of feature alignments are not sufficient to learn domain-invariant representations. To address these limitations, we propose a novel augmented feature alignment network (AFAN) which integrates intermediate domain image generation and domain-adversarial training into a unified framework. An intermediate domain image generator is proposed to enhance feature alignments by domain-adversarial training with automatically generated soft domain labels. The synthetic intermediate domain images progressively bridge the domain divergence and augment the annotated source domain training data. A feature pyramid alignment is designed and the corresponding feature discriminator is used to align multi-scale convolutional features of different semantic levels. Last but not least, we introduce a region feature alignment and an instance discriminator to learn domaininvariant features for object proposals. Our approach significantly outperforms the state-of-the-art methods on standard benchmarks for both similar and dissimilar domain adaptations. Further extensive experiments verify the effectiveness of each component and demonstrate that the proposed network can learn domain-invariant representations. Index Terms-Object Detection, Unsupervised Domain Adaptation I. INTRODUCTION O BJECT detection is a fundamental problem in computer vision, and has been extensively studied for decades. Over the past several years, deep learning has achieved remarkable success in this area [1], [2], [3]. To advance the state-of-the-art on large-scale benchmarks [4], [5], most deep learning based approaches require tremendous amount of annotated training data.In real-world applications, the manual annotating such large-scale data is time-consuming and laborintensive. Moreover, it is challenging to deploy a trained deep learning model to new environments, even if the task is the same. This is simply because the performance is negatively impacted by changes in conditions such as image sensors, viewpoints, weather and time of day.There is a non-negligible discrepancy between the distribution of data from the target domain and that from the source domain.

show abstract

“…This simple approach has the advantage of not requiring any annotated data in the target domain. [45] proposes to improve the distillation by adding additional temporal consistency information obtained via object tracking.…”

Section: Related Workmentioning

confidence: 99%

Self-Supervised Small Soccer Player Detection and Tracking

Hurault

Ballester

Haro

2020

Proceedings of the 3rd International Workshop on Multimedia Content Analysis in Sports

View full text Add to dashboard Cite

In a soccer game, the information provided by detecting and tracking brings crucial clues to further analyze and understand some tactical aspects of the game, including individual and team actions. State-of-the-art tracking algorithms achieve impressive results in scenarios on which they have been trained for, but they fail in challenging ones such as soccer games. This is frequently due to the player small relative size and the similar appearance among players of the same team. Although a straightforward solution would be to retrain these models by using a more specific dataset, the lack of such publicly available annotated datasets entails searching for other effective solutions. In this work, we propose a self-supervised pipeline which is able to detect and track low-resolution soccer players under different recording conditions without any need of ground-truth data. Extensive quantitative and qualitative experimental results are presented evaluating its performance. We also present a comparison to several state-of-the-art methods showing that both the proposed detector and the proposed tracker achieve top-tier results, in particular in the presence of small players. Code available at https:// github.com/ samuro95/ Self-Supervised-Small-Soccer-Player-Detection-Tracking.

show abstract

Automatic Adaptation of Object Detectors to New Domains Using Self-Training

Cited by 139 publications

References 55 publications

Tunnel condition assessment via cloud model‐based random forests and self‐training approach

Tunnel condition assessment via cloud model‐based random forests and self‐training approach

AFAN: Augmented Feature Alignment Network for Cross-Domain Object Detection

Self-Supervised Small Soccer Player Detection and Tracking

Contact Info

Product

Resources

About