RESUMO
Underwater image processing has been shown to exhibit significant potential for exploring underwater environments. It has been applied to a wide variety of fields, such as underwater terrain scanning and autonomous underwater vehicles (AUVs)-driven applications, such as image-based underwater object detection. However, underwater images often suffer from degeneration due to attenuation, color distortion, and noise from artificial lighting sources as well as the effects of possibly low-end optical imaging devices. Thus, object detection performance would be degraded accordingly. To tackle this problem, in this article, a lightweight deep underwater object detection network is proposed. The key is to present a deep model for jointly learning color conversion and object detection for underwater images. The image color conversion module aims at transforming color images to the corresponding grayscale images to solve the problem of underwater color absorption to enhance the object detection performance with lower computational complexity. The presented experimental results with our implementation on the Raspberry pi platform have justified the effectiveness of the proposed lightweight jointly learning model for underwater object detection compared with the state-of-the-art approaches.
RESUMO
Rabies is a fatal zoonosis that causes encephalitis in mammals, and vaccination is the most effective method to control and eliminate rabies. Virus-like vesicles (VLVs), which are characterized as infectious, self-propagating membrane-enveloped particles composed of only Semliki Forest virus (SFV) replicase and vesicular stomatitis virus glycoprotein (VSV-G), have been proven safe and efficient as vaccine candidates. However, previous studies showed that VLVs containing rabies virus glycoprotein (RABV-G) grew at relatively low titers in cells, impeding their potential use as a rabies vaccine. In this study, we constructed novel VLVs by transfection of a mutant SFV RNA replicon encoding RABV-G. We found that these VLVs could self-propagate efficiently in cell culture and could evolve to high titers (approximately 108 focus-forming units [FFU]/ml) by extensive passaging 25 times in BHK-21 cells. Furthermore, we found that the evolved amino acid changes in SFV nonstructural protein 1 (nsP1) at positions 470 and 482 was critical for this high-titer phenotype. Remarkably, VLVs could induce robust type I interferon (IFN) expression in BV2 cells and were highly sensitive to IFN-α. We found that direct inoculation of VLVs into the mouse brain caused reduced body weight loss, mortality, and neuroinflammation compared with the RABV vaccine strain. Finally, it could induce increased generation of germinal center (GC) B cells, plasma cells (PCs), and virus-neutralizing antibodies (VNAs), as well as provide protection against virulent RABV challenge in immunized mice. This study demonstrated that VLVs containing RABV-G could proliferate in cells and were highly evolvable, revealing the feasibility of developing an economic, safe, and efficacious rabies vaccine. IMPORTANCE VLVs have been shown to represent a more versatile and superior vaccine platform. In previous studies, VLVs containing the Semliki Forest virus replicase (SFV nsP1 to nsP4) and rabies virus glycoprotein (RABV-G) grew to relatively low titers in cells. In our study, we not only succeeded in generating VLVs that proliferate in cells and stably express RABV-G, but the VLVs that evolved grew to higher titers, reaching 108 FFU/ml. We also found that nucleic acid changes at positions 470 and 482 in nsP1 were vital for this high-titer phenotype. Moreover, the VLVs that evolved in our studies were highly attenuated in mice, induced potent immunity, and protected mice from lethal RABV infection. Collectively, our study showed that high titers of VLVs containing RABV-G were achieved, demonstrating that these VLVs could be an economical, safe, and efficacious rabies vaccine candidate.
