We present a novel method to train machine learning algorithms to estimate scene depths from a single image, by using the information provided by a camera's aperture as supervision. Prior works use a depth sensor&...
详细信息
ISBN:
(纸本)9781538664209
We present a novel method to train machine learning algorithms to estimate scene depths from a single image, by using the information provided by a camera's aperture as supervision. Prior works use a depth sensor's outputs or images of the same scene from alternate viewpoints as supervision, while our method instead uses images from the same viewpoint taken with a varying camera aperture. To enable learning algorithms to use aperture effects as supervision, we introduce two differentiable aperture rendering functions that use the input image and predicted depths to simulate the depth-of-field effects caused by real camera apertures. We train a monocular depth estimation network end-to-end to predict the scene depths that best explain these finite aperture images as defocus-blurred renderings of the input all-in-focus image.
The availability of affordable 3D full body reconstruction systems has given rise to free-viewpoint video (FVV) of human shapes. Most existing solutions produce temporally uncorrelated point clouds or meshes with unkn...
详细信息
ISBN:
(纸本)9781538664209
The availability of affordable 3D full body reconstruction systems has given rise to free-viewpoint video (FVV) of human shapes. Most existing solutions produce temporally uncorrelated point clouds or meshes with unknown point/vertex correspondences. Individually compressing each frame is ineffective and still yields to ultra-large data sizes. We present an end-to-end deep learning scheme to establish dense shape correspondences and subsequently compress the data. Our approach uses sparse set of "panoramic" depth maps or PDMs, each emulating an inward-viewing concentric mosaics (CM)[45]. We then develop a learning-based technique to learn pixel-wise feature descriptors on PDMs. The results are fed into an autoencoder-based network for compression. Comprehensive experiments demonstrate our solution is robust and effective on both public and our newly captured datasets.
With the recent advancements in multimedia and computergraphics technology, virtual avatar concerts have become increasingly popular among both well-known celebrities and subculture-based virtual YouTubers, attractin...
详细信息
ISBN:
(纸本)9798400701085
With the recent advancements in multimedia and computergraphics technology, virtual avatar concerts have become increasingly popular among both well-known celebrities and subculture-based virtual YouTubers, attracting large audiences. Although this new form of performance art can be delivered through various platforms that use different viewing formats, research on the components of enjoying virtual avatar concerts and their effects on the audience experience is lacking. This study aims to explore the effects of different viewing formats and song genres on sickness, presence, and immersion and identify preferences for enjoying virtual avatar concerts. We first analyzed data from 64 virtual avatar concert cases to identify the most commonly used viewing media, methods, and song genres. Based on our analysis, we designed and conducted a user experiment with four virtual avatar concert scenarios. The experiment compared two viewing methods (audience's point of view and director's edited view) and two song genres (dance and ballad) for 2D display and a head-mounted display. The results show that the audience's point of view and a dance song can be advantageous in increasing the immersive experience irrespective of the viewing medium. Furthermore, the participants show different expectations regarding performances according to song genres. We discuss these findings and propose guidelines to help design and conduct virtual avatar concerts effectively.
Multi-touch gesture interaction is one of the main features that differentiate the current smart mobile era from the desktop-computing era. Heuristic evaluation is a popular usability evaluation method, mainly due to ...
详细信息
The proceedings contain 18 papers. The topics discussed include: towards a generic framework for multi-party dialogue with virtual humans;increasing the feeling of social presence by incorporating realistic interactio...
ISBN:
(纸本)9781450363761
The proceedings contain 18 papers. The topics discussed include: towards a generic framework for multi-party dialogue with virtual humans;increasing the feeling of social presence by incorporating realistic interactions in multi-party VR;learning to communicate via supervised attentional message processing;socio-cultural effects of virtual counseling interviewers as mediated by smartphone video conferencing;an image representation for the 3D face synthesis;and saliency detection by aggregating complementary background template with foreground information.
Rendering of large scenes in external memory is one of the most important problems of computergraphics, which is used in such areas as CAD/CAM/CAE, geoinformatics, project management, scientific visualization, virtua...
详细信息
Car/ride sharing concepts are getting more and more popular, mainly due to the more economical and environment-friendly nature. Most of the popular car/ride service providers offer specialized mobile apps so that user...
详细信息
The modern level of computergraphics programs (CGP) has long been able to solve most design, architecture, and lighting design problems. However, the use of CGP in teaching students is still underestimated. The exist...
详细信息
The proceedings contain 14 papers. The topics discussed include: reducing error aversion to support novice-to-expert transitions with FastTap;tactile and kinesthetic feedbacks improve distance perception in virtual re...
ISBN:
(纸本)9781450370264
The proceedings contain 14 papers. The topics discussed include: reducing error aversion to support novice-to-expert transitions with FastTap;tactile and kinesthetic feedbacks improve distance perception in virtual reality;which tangible control for which visual task?;multi-evaluation of pointing and application to dyspraxia evaluation;mulsebox: a new multisensorial interaction device;training road crossing skills for young people with intellectual disabilities using virtual reality - a feasibility study;holding patterns: detecting handedness with a moving smartphone at pickup;and vibrotactile cues for conveying directional information during blind exploration of digital graphics.
The high processing capabilities of current smartphones and the availability of wearable Virtual Reality (VR) toolkits make it possible for normal users to use VR environments on go for gaming or learning. However, us...
详细信息
暂无评论