arXiv subjects

Kaifeng Zhang

Publications and source records attributed to Kaifeng Zhang.

39 records · Page 3Linked to original sources

arXiv2023

Self-Supervised Geometric Correspondence for Category-Level 6D Object Pose Estimation in the Wild

While 6D object pose estimation has wide applications across computer vision and robotics, it remains far from being solved due to the lack of annotations. The problem becomes even more challenging when moving to category-level 6D pose, which requires generalization to unseen instances. Current approaches are restricted by leveraging annotations from simulation or collected from humans. In this paper, we overcome this barrier by introducing a self-supervised learning approach trained directly on large-scale real-world object videos for category-level 6D pose estimation in the wild. Our framework reconstructs the canonical 3D shape of an object category and learns dense correspondences between input images and the canonical shape via surface embedding. For training, we propose novel geometrical cycle-consistency losses which construct cycles across 2D-3D spaces, across different instances and different time steps. The learned correspondence can be applied for 6D pose estimation and other downstream tasks such as keypoint transfer. Surprisingly, our method, without any human annotations or simulators, can achieve on-par or even better performance than previous supervised or semi-supervised methods on in-the-wild images. Our project page is: https://kywind.github.io/self-pose .

cs.CV↗

arXiv2022

Semantic-Aware Fine-Grained Correspondence

Establishing visual correspondence across images is a challenging and essential task. Recently, an influx of self-supervised methods have been proposed to better learn representations for visual correspondence. However, we find that these methods often fail to leverage semantic information and over-rely on the matching of low-level features. In contrast, human vision is capable of distinguishing between distinct objects as a pretext to tracking. Inspired by this paradigm, we propose to learn semantic-aware fine-grained correspondence. Firstly, we demonstrate that semantic correspondence is implicitly available through a rich set of image-level self-supervised methods. We further design a pixel-level self-supervised learning objective which specifically targets fine-grained correspondence. For downstream tasks, we fuse these two kinds of complementary correspondence representations together, demonstrating that they boost performance synergistically. Our method surpasses previous state-of-the-art self-supervised methods using convolutional networks on a variety of visual correspondence tasks, including video object segmentation, human pose tracking, and human part tracking.

cs.CV↗

arXiv2020

Vulnerability Assessment of Load Frequency Control Considering Cyber Security

Security is one of the biggest concern in power system operation. Recently, the emerging cyber security threats to operational functions of power systems arouse high public attention, and cybersecurity vulnerability thus become an emerging topic to evaluate compromised operational performance under cyber attack. In this paper, vulnerability of cyber security of load frequency control (LFC) system, which is the key component in energy manage system (EMS), is assessed by exploiting the system response to attacks on LFC variables/parameters. Two types of attacks: 1) injection attack and 2) scale attack are considered for evaluation. Two evaluation criteria reflecting the damage on system stability and power generation are used to quantify system loss under cyber attacks. Through a sensitivity-based method and attack tree models, the vulnerability of different LFC components is ranked. In addition, a post-intrusion cyber attack detection scheme is proposed. Classification-based schemes using typical classification algorithms are studied and compared to identify different attack scenarios.

eess.SY↗