Recent development in sound technologies has enabled the realistic replay of real-life sounds. Thanks to these technologies, we can experience a virtual real sound environment. However, there are other types of sound technologies that enhance reality, such as acoustic filters, sound effects, and background music. They are quite effective if carefully prepared, but they also alter the sound itself. Consequently, sound is simultaneously used to reconstruct realistic environments and to enhance emotions, which are actually incompatible functions.With this background, we focused on using tactile modality to enhance emotions and propose a method that enhances the sound experience by a combination of sound and skin sensation to the pinna (earlobe). In this paper, we evaluate the effectiveness of this method.

Computer games provide users with a mental stimulation that the real world cannot. Especially, horror games are a popular category. Current horror games can provide the user with a visible ghost and the stereo background sound to thrill the user. Inspired by obstacle sense -the ability of blind people localizing themselves only with hearing, a novel method to augment the sense of existence in the game background sound is proposed in this paper. We found that an effective sense can be created by decreasing high frequency component and increasing low frequency component simultaneously.

No abstract

No abstract

The major approaches of transfer learning in computer vision have tried to adapt the source domain to the target domain one-to-one. However, this scenario is difficult to apply to real applications such as video surveillance systems. As those systems have many cameras installed at each location regarded as source domains, it is difficult to identify the proper source domain. In this paper, we introduce a new transfer learning scenario that has various source domains and one target domain, assuming video surveillance system integration. Also, we propose a novel method for automatically producing a high accuracy model by fusing models trained at various source domains. In particular, we show how to apply a gating network to fuse source domains for object detection tasks, which is a new approach. We demonstrate the effectiveness of our method through experiments on traffic surveillance datasets.