Strong-Weak Distribution Alignment for Adaptive Object Detection

Saito, Kuniaki; Ushiku, Yoshitaka; Harada, Tatsuya; Saenko, Kate

doi:10.1109/cvpr.2019.00712

Cited by 654 publications

(867 citation statements)

References 38 publications

Supporting

Mentioning

821

Contrasting

Order By: Relevance

“…Experiments have shown that aligning the feature distributions of intermediate layers can also alleviate covariate shift and achieve better domain adaptation. Furthermore, our model also follows the conclusion in [21] that local alignment should be stronger than global alignment. Because during the backpropagation, the lower feature extractors in Faster R-CNN are getting the reversal gradient from all subsequent domain classifiers, which means it should maintain stronger ability of feature alignment to deceive more domain classifiers.…”

Section: Introductionsupporting

confidence: 77%

“…In addition, we find the performance can be further improved by increasing the number of domain classifiers from 3 to 6. Method G I CTX L Car AP Source(Supervised) 34.3 DA Model * [2] 39.0 DA Model [2] 39.4 SW-DA [21] 40.1 SW-DA(γ = 3) [21] 42.3 SC-DA(Type3) [29] 43…”

Section: Synthetic Data Adaptationmentioning

confidence: 99%

“…Domain Adaptive (DA) Faster R-CNN [2] extends this idea to object detection, matching features in the image stage and instance stage. In a more recent work, the authors demonstrate that global matching in DA Faster R-CNN may only address small domain shifts, but the detection accuracy related to large domain shifts may decrease [21]. Therefore, they propose a Strong-Weak DA Faster R-CNN that combines weak global alignment with strong local alignment.…”

Section: Introductionmentioning

confidence: 99%

“…That is, Faster R-CNN can easily deceive the domain classifier, so that the feature alignment is highly possible to be ineffective. Inspired by the Strong-Weak DA Faster R-CNN [21], we propose a Multi-level Domain Adaptive Faster R-CNN. Our model has two advantages: First, we use different domain classifiers to supervise the feature alignments from multiple scales; Second, more domain classifiers enhance the model's discriminating ability and optimize overall adversarial training.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Multi-Level Domain Adaptive Learning for Cross-Domain Detection

Xie

Wang

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)

101

View full text Add to dashboard Cite

In recent years, object detection has shown impressive results using supervised deep learning, but it remains challenging in a cross-domain environment. The variations of illumination, style, scale, and appearance in different domains can seriously affect the performance of detection models. Previous works use adversarial training to align global features across the domain shift and to achieve image information transfer. However, such methods do not effectively match the distribution of local features, resulting in limited improvement in cross-domain object detection. To solve this problem, we propose a multi-level domain adaptive model to simultaneously align the distributions of local-level features and global-level features. We evaluate our method with multiple experiments, including adverse weather adaptation, synthetic data adaptation, and cross camera adaptation. In most object categories, the proposed method achieves superior performance against stateof-the-art techniques, which demonstrates the effectiveness and robustness of our method.

show abstract

Section: Introductionsupporting

confidence: 77%

Section: Synthetic Data Adaptationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Multi-Level Domain Adaptive Learning for Cross-Domain Detection

Xie

Wang

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)

101

View full text Add to dashboard Cite

show abstract

“…Hence, an effective model that can adapt object detectors into a new domain without labels, i.e., unsupervised domain adaptation, is highly desirable. However, only a few works [24,25] are proposed to address the unsupervised domain adaptation problem for object detection.…”

Section: Goals and Challengesmentioning

confidence: 99%

On the robustness of object and face detection: false positives, attacks and adaptability

Siqi¹

View full text Add to dashboard Cite

Object detection is to recognize and locate objects in images, which is one of the fundamental problems in computer vision. It has been developed rapidly in the past decades and has a wide range of applications, e.g., surveillance systems, autonomous vehicles, medical imaging diagnosis, and industrial automation systems. Since the technique of object detection has been widely adopted, the lack of robustness of object detectors can be problematic, which may lead to unpredictable safety and financial losses. The main goal of this thesis is to study the robustness of object detectors and develop robust detectors. Robust object detectors can be viewed as the detectors that have the ability to maintain high accuracy when encountering the challenging conditions that may degrade the performance. To develop robust object detectors, we study the robustness of object detection by understanding when and where the detectors may fail. In this thesis, we focus on the failure cases in three aspects: false positives, attacks, and domain adaptability. False positives are the predictions that do not match the ground truth bounding boxes. The robustness to false positives is crucial for object detection, particularly for face detection, as face detection problem is the very first step for facial analysis tasks. The existence of false positives will waste the computations and harm the accuracy of the subsequent processes. To make the current detectors more robust, this thesis focuses on reducing the number of false positives with the detection rate well-maintained. To achieve this, we first explore different post-processing false positive classifiers to remove false positives. We further study the properties of true and false positives among face detectors and propose a framework that cascades two off-the-shelf detectors. The cascaded detector demonstrated its effectiveness and efficiency on false positive removal. Attacks are one of the biggest threats of a robust detection system, especially for surveillance systems. For example, criminals can apply attacks to the surveillance system for preventing themselves from being detected. Adversarial attacks are the attacks that can change the neural network output significantly by adding tiny, imperceptible perturbations onto the image. We examine how robust the current detectors are against adversarial attacks and provide theoretical explanations for why existing adversarial attack methods fail. We successfully perform adversarial attacks on deep learning based detection networks. Due to the variations in shape and appearance, lighting conditions and backgrounds, a model trained on the source data might not perform well on the target data, which is often known as domain discrepancy. Supervised domain adaptation methods require a large number of annotated bounding boxes for object detection, which is time-consuming and expensive. Hence, an effective model that can adapt object detectors into a new domain without labels is highly desirable. In this thesis, we propose an adaptation method t...

show abstract

A Knowledge Transfer Method for Unsupervised Pose Keypoint Detection Based on Domain Adaptation and CAD Models

Kong

Delong

2023

Advanced Intelligent Systems

View full text Add to dashboard Cite

Vision‐based pose estimation is a basic task in many industrial fields such as bin‐picking, autonomous assembly, and augmented reality. One of the most commonly used pose estimation methods first detects the 2D pose keypoints in the input image and then calculates the 6D pose using a pose solver. Recently, deep learning is widely used in pose keypoint detection and performs excellent accuracy and adaptability. However, its over‐reliance on sufficient and high‐quality samples and supervision is prominent, particularly in the industrial field, leading to high data cost. Based on domain adaptation and computer‐aided‐design (CAD) models, herein, a virtual‐to‐real knowledge transfer method for pose keypoint detection to reduce the data cost of deep learning is proposed. To address the disorder of knowledge flow, a viewpoint‐driven feature alignment strategy is proposed to simultaneously eliminate interdomain differences and preserve intradomain differences. The shape invariance of rigid objects is then introduced as constraints to address the large assumption space problem in the regressive domain adaptation. The multidimensional experimental results demonstrate the superiority of the method. Without real annotations, the normalized pixel error of keypoint detection is reported as 0.033, and the proportion of pixel errors lower than 0.05 is up to 92.77%.

show abstract

Strong-Weak Distribution Alignment for Adaptive Object Detection

Cited by 654 publications

References 38 publications

Multi-Level Domain Adaptive Learning for Cross-Domain Detection

Multi-Level Domain Adaptive Learning for Cross-Domain Detection

On the robustness of object and face detection: false positives, attacks and adaptability

A Knowledge Transfer Method for Unsupervised Pose Keypoint Detection Based on Domain Adaptation and CAD Models

Contact Info

Product

Resources

About