Search

Multi-viewpoint image stitching aims to stitch images taken from different viewpoints into pictures with a broader field of view. The stitched images are subject to artifacts, geometric distortion, and blur distortion due to the mismatch of feature points, inaccurate homography estimation, and improper fusion of the unstitched images. Deep learning has recently been increasingly applied to multi-viewpoint image stitching to overcome these problems. However, there has thus far been little related research to summarize the different deep learning techniques used for multi-viewpoint image stitching. Therefore, this review aims to explore the application of deep learning to multi-viewpoint image stitching. To better illustrate this topic, we first summarize the acquisition methods for multi-viewpoint images and the main challenges of image stitching. After which, deep learning techniques for multi-view image stitching with a single camera are sorted out. Subsequently, deep learning techniques for multi-view image stitching with camera arrays, including parallel-view multi-view image stitching and cross-view multi-view image stitching, are presented. Next, we summarize image stitching datasets, evaluation metrics, and experimental data of several leading stitching algorithms on public datasets. Finally, we discuss potential issues and future work on image stitching with multi-viewpoint images. [ABSTRACT FROM AUTHOR]

Due to light absorption and scattering underwater images suffer from low contrast, color distortion, blurred details, and uneven illumination, which affect underwater vision tasks and research. Therefore, underwater image enhancement is of great significance in vision applications. In contrast to existing methods for specific underwater environments or reliance on paired datasets, this study proposes an underwater multiscene generative adversarial network (UMGAN) to enhance underwater images. The network implements unpaired image-to-image translation between the underwater turbid domain and the underwater clear domain. It has a great enhancement impact on several underwater image types. Feedback mechanisms and a noise reduction network are designed to optimize the generator and address the issue of noise and artifacts in GAN-produced images. Furthermore, a global–local discriminator is employed to improve the overall image while adaptively modifying the local region image effect. It resolves the issue of over- and underenhancement in local regions. The reliance on paired training data is eliminated through a cycle consistency network structure. UMGAN performs satisfactorily on various types of data when compared quantitatively and qualitatively to other state-of-the-art algorithms. It has strong robustness and can be applied to various enhancement tasks in different scenes. [ABSTRACT FROM AUTHOR]

Similar to traditional luxuries, affordable luxuries enjoy a high level of perceived product quality and perceived social status, but the effectiveness of price promotions of purchasing affordable luxury products is different from that of traditional luxuries. In order to further investigate the purchases of affordable luxuries, we used event-related potential (ERP) technology to reveal the formation of the purchase intention toward affordable luxuries at original prices (high or low) and current prices (discounted or non-discounted). Compared with the high-priced affordable luxury without a price promotion and the low-priced affordable luxury with a price promotion, consumers showed a stronger intention toward the high-priced affordable luxury with a price promotion, by weighing up three factors, perceived product quality, perceived social status, and perceived monetary saving at the behavioral level. A shorter reaction time emerged in the price promotion condition than in the absent price promotion condition when the original price was low. At the neural level, a decrease in N2 amplitude was found in the high original price and discounted current price condition than the high original price and non-discounted current price condition and the low original price and discounted current price condition, respectively, suggesting that the price information of the latter two conditions might not be the expectation information of subjects, and thus, the enhanced conflict is produced. The high-priced affordable luxury product without a price discount evoked a more positive LPP amplitude than the high-priced affordable luxury product with price promotions or than the low-priced affordable luxury item without price promotions, demonstrating that participants could regard the former as an evaluative inconsistent condition and the latter as evaluative consistent conditions. These results are helpful to better understand the effects of price promotions on purchasing affordable luxury products at different original prices. [ABSTRACT FROM AUTHOR]

There is a connection between the frontal negative slow wave (FNSW) and the arousal inhibition in the hedonic purchase context. To calculate the FNSW (400–800 ms), event-related potentials (ERPs) method was applied to depict the neural substrates on prudent and impulsive consumers' behaviors within various states of promotion. Promotion types include the pure price promotion and the mixed promotion (a mixture of a charitable donation and a discount). Behaviorally, consumers response more quickly in the pure price promotion condition and they express a preference for the mixed promotion. More importantly, a larger FNSW emerged in the impulsive consumers than the prudent, suggesting that the former might tend to control their eagerness to consume hedonic items. Compared with the price promotion as the worse option, the mixed promotion as the better option caused more perceptual conflict, leading to an increase in N2 amplitude. It suggests that consumers incline to reject the worse offers. These results also reveal that people primarily have to search negative promotion information by their insight and subsequently impulsive consumers inhibit the responses to the promotion information. The method of ERPs and FNSW should be helpful for marketing researchers and professionals on hedonic consumption and sales promotion. [ABSTRACT FROM AUTHOR]

Due to consuming hedonic products unnecessary to basic well-being, consumers need justifications for pleasure. However, different justifications have differential influences in promoting hedonic purchases, such as price and quantity promotions (PP and QP), the difference being that the latter requires purchasing additional units to get the same discount as the former. In the present study, even-related potentials (ERPs) was applied to reveal the timing of brain activities to further understand how promotion information consisting of promotion type (PP and QP) and discount depth, deep and shallow discounts (DD and SD) on hedonic products was processed. Behaviorally, consumers were more willing to purchase items in PP and DD conditions than QP and SD conditions, respectively, and spent more time making final purchase decisions in QP and DD condition or PP and SD condition compared to PP and DD condition. Neurophysiologically, DD automatically recruited more attentional resources than SD and led to a higher P2 amplitude. QP and DD condition or PP and SD condition evoked a larger N2 amplitude and enhanced perceptual conflict compared to PP and DD condition. During late stage, PP and DD elicited a more positive LPP amplitude in contrast to QP and SD, respectively, indicating that people have stronger purchase intention and positive affect in PP and DD contexts. These findings provided evidence for the differential influences between PP and QP and what ultimately made consumers buy hedonic products or not. [ABSTRACT FROM AUTHOR]