no code implementations • 4 Mar 2020 • Xingyu Chen, Yue Lu, Zhengxing Wu, Junzhi Yu, Li Wen
According to our analysis, five key discoveries are reported: 1) Domain quality has an ignorable effect on within-domain convolutional representation and detection accuracy; 2) low-quality domain leads to higher generalization ability in cross-domain detection; 3) low-quality domain can hardly be well learned in a domain-mixed learning process; 4) degrading recall efficiency, restoration cannot improve within-domain detection accuracy; 5) visual restoration is beneficial to detection in the wild by reducing the domain shift between training data and real-world scenes.
no code implementations • 22 Dec 2019 • Xingyu Chen, Zhengxing Wu, Junzhi Yu, Li Wen
From a robotic perspective, the importance of recall continuity and localization stability is equal to that of accuracy, but the AP is insufficient to reflect detectors' performance across time.
1 code implementation • 23 Jul 2018 • Xingyu Chen, Junzhi Yu, Shihan Kong, Zhengxing Wu, Li Wen
As for temporal detection in videos, temporal refinement networks (TRNet) and temporal dual refinement networks (TDRNet) are developed by propagating the refinement information across time.
1 code implementation • 3 Dec 2017 • Xingyu Chen, Junzhi Yu, Shihan Kong, Zhengxing Wu, Xi Fang, Li Wen
More specifically, an underwater index is investigated to describe underwater properties, and a loss function based on the underwater index is designed to train the critic branch for underwater noise suppression.