Search results
Dr. Qing Guo is currently a senior research scientist and principal investigator (PI) at the Center for Frontier AI Research (CFAR), A*STAR in Singapore. He is also an adjunct assistant professor at the National University of Singapore (NUS) .
Introduction. In this paper, we propose a generalized RACM by defining frequency boundary energy (FBE) in the frequency domain. Image segmentation is defined as finding a contour that minimizes the FBE outside and inside the contour. Thus, the segmentation result is decided by a frequency filter that produces FBE.
We then formulate the online selection of these weight maps as a decision making problem by a Markov Decision Process (MDP), where the learning of weight map selection is equivalent to policy learning of the MDP that is solved by a reinforcement learning strategy.
DeepFake detection, heartbeat rhythm, remote photoplethysmog-raphy (PPG), dual-spatial-temporal attention, face forensics. ACM Reference Format: Hua Qi1∗, Qing Guo2∗, Felix Juefei-Xu3, Xiaofei Xie2, Lei Ma1† and Wei Feng4, Yang Liu2, Jianjun Zhao1. 2020. DeepRhythm: Exposing DeepFakes with Attentional Visual Heartbeat Rhythms.
Basic pipelines of our DSiam network (orange line) and the SiamFC (black dashed line). f l (⋅) represents a CNN to extract the deep feature at l th layer. We add the target appearance variation (V t - 1 l) and background suppression (W t - 1 l) transformations for two branches respectively.
Qing Guo, Wei Feng*, Ce Zhou, Bin Wu. University, Tianjin, Chinaftsingqguo, wfeng, zhouce,binwug@tju.edu.cnABSTRACTCompressive random projection is a powerful appearance model to derive effective Haar-like features from.
In this paper, we propose background-suppressed correlation filters~ (BSCF), a better CF tracking scheme, which can significantly improve the reliability and accuracy of CF trackers, without harming their beyond real-time speed. Specifically, we present a unified BSCF object function.
Visual tracking aims to track an arbitrary temporally-changing object, with the target being only specified at the first frame. Since potential changes of the object and its context are basically unknown and constantly happen, ∗Corresponding author. Tel: (+86)-22-27406538.
Results on OTB-2015. Precision plots (left) and success plots (right) showing a comparison with state-of-the-art methods on OTB-2015. The legend contains the average distance precision score at 20 pixels and the AUC score for each tracker.
Shuifa Sun **, Qing Guo *, Fangmin Dong, Bangjun Lei Institute of Intelligent Vision and Image Information, China Three Gorges University, Yichang 443002, China ABSTRACT Â In this paper, a real-time visual tracking system that delivers superior performance under difficult situations is proposed. The system is based on Histogram of Oriented ...
