Showing papers in "Displays in 2022"

TL;DR: In this article , a parallel bi-branch model (Trans-CNN Net) based on Transformer module and Convolutional Neural Network module is proposed by making full use of the local feature extraction capability of CNN and the global feature extraction advantage of Transformer.

TL;DR: In this article , a comparative study of the state-of-the-art Electrical Resistivity Tomography applications in the context of underground imaging and object detection is presented, which includes underground utilities, underground cavities, old and ancient structures, fault zones and potential construction sites, hydrological features and bedrock, concrete and other objects.

TL;DR: In this paper , a comprehensive look at the recent works using GANs for generating visual arts, music, and literary text is presented. And some of the key challenges in GAN-based art generation are highlighted along with recommendations for future work.

TL;DR: Zhang et al. as discussed by the authors proposed a two-stage HGR system, which first performs accurate segmentation to segment the hand from the background and then uses double-channel CNNs to improve the recognition performance.

TL;DR: Li et al. as discussed by the authors adopt the teacher-student framework to generate pseudo-labels from unlabeled training data, and use a label filtering method to improve the pseudo label quality.

TL;DR: In this article , the influence of dynamic images and traditional static images on user perception in web interface visualizations of products was discussed, and the results of eye movement experiment show that the visual cognitive effect was better for the dynamic images than the static image, and efficiency of visual search was improved.

TL;DR: In this paper , a review of explainable Artificial Intelligence (XAI) techniques applied to intelligent systems to justify their decisions offers a possible solution to the lack of system interpretability is a major obstacle to the wider adoption of AI in safety-critical applications.

TL;DR: In this article , a multi-level cross-modal alignment (MCA) module is proposed to align the image scene graph with the sentence scene graph at different levels, which can distill the redundant information of the image scenes graph according to the sentence scenes graph, and provide commonsense knowledge for the decoder.

TL;DR: Wang et al. as discussed by the authors proposed a deep learning model for mask detection and social distance measurement, which achieved a comparatively high mean average precision of 92.4% for YOLOR.

TL;DR: In this article , the Virtual Special Issue: 3D Data Computation and Visualization (VCDV) is presented. ǫ 3D data computation and visualization was handled in 2021.

TL;DR: DeepRPN-BIQA as discussed by the authors proposes a region proposal network (RPN) for blind natural-scene and screen-content image quality assessment, which computes visual saliency using RPN to extract important regions having a high contribution towards the image quality.

TL;DR: DeepRPN-BIQA as mentioned in this paper proposes a region proposal network (RPN) for blind natural-scene and screen-content image quality assessment, which computes visual saliency using RPN to extract important regions having a high contribution towards the image quality.

TL;DR: In this article, a study focused on the affective characteristics of product shapes and colours and involved two experiments: (1) an emotional characteristic experiment on the imagery of product shape, and (2) a metallic product colour design experiment.

TL;DR: Wang et al. as mentioned in this paper proposed a DCT based frequency attention network (DCT-FANet) to distinguish different frequency of LR images, and enhance the high-frequency information adaptively.

TL;DR: In this article , the intermediate reference features are synthesized in gradient domain, structural tensor domain, and global perception domain, respectively, and a single quality measure is estimated to reflect the visual quality of the fused image from a specific perspective.

TL;DR: In this article , the authors investigated whether severe symptoms of visually induced motion sickness (VIMS) can occur in augmented reality (AR) optical see-through applications, in which the real world is enhanced by virtual objects.

TL;DR: In this article , an encoder based modified Generative Adversarial Network (GAN) architecture is proposed for land use classification which can yield better results with less amount of training samples. And the validation of the proposed model is carried out on UC Merced and AID land use datasets.

TL;DR: In this paper , a study focused on the affective characteristics of product shapes and colours and involved two experiments: (1) an emotional characteristic experiment on the imagery of product shape, (2) a metallic product colour design experiment.

TL;DR: In this article , a Pseudo-CNN (P-CNN) is proposed to remove Gaussian-Impulse noise with blind (unknown) noise densities in digital images, which has two phases; in phase 1, the noisy image is preprocessed with a novel Pseudo CNN having no tunable parameters.

TL;DR: Wang et al. as mentioned in this paper developed an AI-assisted diagnosis method that employs Faster-RCNN to predict the number and locations of caries lesions based on periapical films.

TL;DR: Wang et al. as mentioned in this paper proposed a simple yet effective convolution neural network (CNN)-based framework to enhance underwater images, which consists of two stages, CNN-based enhancement and YUV-based post-processing.

TL;DR: In this article , the trapezoidal chip structure, surface grating, and surface coarsening are simulated in order to obtain their contributions to enhance the light extraction efficiency of micro-LEDs.

TL;DR: Wang et al. as mentioned in this paper proposed a multi-level U-Net network (MUN) for image super-resolution reconstruction, which used dual-branch convolutional layers with different kernels.

TL;DR: Wang et al. as discussed by the authors mainly studied the key technologies such as 3D image mosaic, surface reconstruction and model simplification, studies the 2D image filtering, and analyzes the transmission of 3D technology and digital media technology.

TL;DR: Zhang et al. as discussed by the authors proposed a zero-shot action recognition model which jointly captures object relations of one static frame and models temporal motion patterns of adjacent frames, and a visual prior-knowledge learning method is devised to introduce visual commonsense knowledge for the accurate identification of unseen action classes.

TL;DR: In this article , LiSrP3O9:Sm3+ phosphors were successfully prepared by the solution combustion method and the powder X-ray diffraction analysis confirmed a single phase for the samples with a crystallite size ranging from 66.9 nm to 72 ± 0.5 nm.

TL;DR: In this paper, the authors found that perceived visual motion (illusory self-motion) is sufficient to elicit vection in the absence of any stimulated visual motion, with rotating illusions continuously perceived as moving more compared to expanding motion illusions, which could be related to missing stereoscopic motion-in-depth cues.

TL;DR: In this paper , a cross-attention redistribution (CAReD) module is proposed to adaptively integrate support features into query features, effectively removing harmful support features and enhancing the regional features of novel categories.

TL;DR: In this article , the most suitable colors for interactive webpages were royal blue (R65, G105, B225), slate blue (G90, B205, and dark blue), while the results can serve as references for researchers and website designers.

TL;DR: In this article , a novel hyperparameter optimization based deep learning model for automated bone age assessment and classification (HPTDL-BAAC) is presented. And the proposed model involves data normalization for pre-processing.