Active Recurrence of Lighting Condition for Fine-Grained Change Detection

Zhang, Qian; Feng, Wei; Wan, Liang; Tian, Fei-Peng; Tan, Ping

doi:10.24963/ijcai.2018/690

Cited by 12 publications

(5 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Active camera relocalization, a critical step in the finegrained change detection task (Feng et al, 2015), is proposed to overcome the image distortion caused by the virtual image alignment . Unlike Active recurrence of lighting (ALR), which is used to actively reproduce the lighting conditions of the rephotograph (Zhang et al, 2018), ACR employs a mechanical platform to physically adjust the camera itself to return to the position in which the reference image was taken. So far, ACR has been widely used in the study of history (Wells II, 2012), monitoring of the natural environment (Guggenheim et al, 2006), or high-value scenes (e.g., cultural heritage) (Feng et al, 2015).…”

Section: Active Camera Relocalizationmentioning

confidence: 99%

Illumination-Invariant Active Camera Relocalization for Fine-Grained Change Detection in the Wild

Li¹,

Feng²,

Zhang³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

Active camera relocalization (ACR) is a new problem in computer vision that significantly reduces the false alarm caused by image distortions due to camera pose misalignment in fine-grained change detection (FGCD). Despite the fruitful achievements that ACR can support, it still remains a challenging problem caused by the unstable results of relative pose estimation, especially for outdoor scenes, where the lighting condition is out of control, i.e., the twice observations may have highly varied illuminations. This paper studies an illumination-invariant active camera relocalization method, it improves both in relative pose estimation and scale estimation. We use plane segments as an intermediate representation to facilitate feature matching, thus further boosting pose estimation robustness and reliability under lighting variances. Moreover, we construct a linear system to obtain the absolute scale in each ACR iteration by minimizing the image warping error, thus, significantly reduce the time consume of ACR process, it is nearly 1.6 times faster than the stateof-the-art ACR strategy. Our work greatly expands the feasibility of real-world fine-grained change monitoring

show abstract

Section: Active Camera Relocalizationmentioning

confidence: 99%

Illumination-Invariant Active Camera Relocalization for Fine-Grained Change Detection in the Wild

Li¹,

Feng²,

Zhang³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…First, we use the commonly-used point light source (PLS) to simulate the estimated adversarial lighting. Second, following the state-ofthe-art active lighting recurrence (ALR) method [45], we can physically adjust the position of the PLS by a robotic arm and produce the real lighting condition that is the same to the estimated adversarial one. Fig.…”

Section: Physical Ara (Phy-ara)mentioning

confidence: 99%

“…In contrast to existing ALR method [45] that depends on the parallel lighting model, our adversarial relighting is based on the spherical harmonics form that is a more general lighting representation and cannot be processed via the ALR directly. To fill the gap, we change the generation manner of lighting map in the ALR and enable it to support our experiment.…”

Section: Physical Ara (Phy-ara)mentioning

confidence: 99%

“…4, we can get the instant navigation feedback from M and M t . After that, we employ the incremental adjustment strategy in [45] to actively tune a robotic arm and update L t to match L. Finally, we get the physical adversarial light (i.e., Lphy ) under which we take a new image as the physical adversarial relighting image (i.e., Îphy ). As shown in Fig.…”

Section: Physical Ara (Phy-ara)mentioning

confidence: 99%

See 1 more Smart Citation

Adversarial Relighting against Face Recognition

Gao¹,

Gao²,

Zhang³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Deep face recognition (FR) has achieved significantly high accuracy on several challenging datasets and fosters successful real-world applications, even showing high robustness to the illumination variation that is usually regarded as a main threat to the FR system. However, in the real world, illumination variation caused by diverse lighting conditions cannot be fully covered by the limited face dataset. In this paper, we study the threat of lighting against FR from a new angle, i.e., adversarial attack, and identify a new task, i.e., adversarial relighting. Given a face image, adversarial relighting aims to produce a naturally relighted counterpart while fooling the state-of-the-art deep FR methods. To this end, we first propose the physical modelbased adversarial relighting attack (ARA) denoted as albedoquotient-based adversarial relighting attack (AQ-ARA). It generates natural adversarial light under the physical lighting model and guidance of FR systems and synthesizes adversarially relighted face images. Moreover, we propose the auto-predictive adversarial relighting attack (AP-ARA) by training an adversarial relighting network (ARNet) to automatically predict the adversarial light in a one-step manner according to different input faces, allowing efficiency-sensitive applications . More importantly, we propose to transfer the above digital attacks to physical ARA (Phy-ARA) through a precise relighting device, making the estimated adversarial lighting condition reproducible in the real world. We validate our methods on three state-of-the-art deep FR methods, i.e., FaceNet, ArcFace, and CosFace, on two public datasets. The extensive and insightful results demonstrate our work can generate realistic adversarial relighted face images fooling FR easily, revealing the threat of specific light directions and strengths.

show abstract

“…4, we can get the instant navigation feedback fromM and M t . After that, we employ the incremental adjustment strategy in [45] to actively tune a robotic arm and update L t to matchL. Finally, we get the physical adversarial light (i.e.,L phy ) under which we take a new image as the physical adversarial relighting image (i.e., I phy ).…”

Section: B Auto-predictive-based Ara (Ap-ara)mentioning

confidence: 99%

A Novel Adversarial Relighting via Face Recognition

Shizutoshi¹

2021

Preprint

View full text Add to dashboard Cite

Deep facial recognition (FR) has reached very high accuracy on various demanding datasets and encourages successful real-world applications, even demonstrating strong tolerance to illumination change, which is commonly viewed as a major danger to FR systems. In the real world, however, illumination variance produced by a variety of lighting situations cannot be adequately captured by the limited facsimile. To this end, we first propose the physical model- based adversarial relighting attack (ARA) denoted as albedo- quotient-based adversarial relighting attack (AQ-ARA). It generates natural adversarial light under the physical lighting model and guidance of FR systems and synthesizes adversarially relighted face images. Moreover, we propose the auto-predictive adversarial relighting attack (AP-ARA) by training an adversarial relighting network (ARNet) to automatically predict the adversarial light in a one-step manner according to different input faces, allowing efficiency-sensitive applications . More importantly, we propose to transfer the above digital attacks to physical ARA (Phy- ARA) through a precise relighting device, making the estimated adversarial lighting condition reproducible in the real world. We validate our methods on three state-of-the-art deep FR methods, i.e., FaceNet, ArcFace, and CosFace, on two public datasets. The extensive and insightful results demonstrate our work can generate realistic adversarial relighted face images fooling FR easily, revealing the threat of specific light directions and strengths.

show abstract

Active Recurrence of Lighting Condition for Fine-Grained Change Detection

Cited by 12 publications

References 34 publications

Illumination-Invariant Active Camera Relocalization for Fine-Grained Change Detection in the Wild

Illumination-Invariant Active Camera Relocalization for Fine-Grained Change Detection in the Wild

Adversarial Relighting against Face Recognition

A Novel Adversarial Relighting via Face Recognition

Contact Info

Product

Resources

About