Titlebook: Advances in Intelligent Information Hiding and Multimedia Signal Processing; Proceedings of the T Jeng-Shyang Pan,Pei-Wei Tsai,Lakhmi C. Ja

連詞

https://doi.org/10.1007/978-3-658-01524-4nt of internet and mobile services, the use of 3D video application from the big screen to mobile devices become an inevitable trend. So an efficient transcoding for 3D videos is necessary. In 3D-HEVC, 3D video is comprised by multiview video and corresponding depth maps. The computational complexit

Mercantile

Becoming a Business Consultant,enables a drone to publish a live video to streaming media server directly. The Qualcomm Snapdragon Flight, a highly integrated board that targets consumer drones applications, is the experimental platform we use. The camera on the board is used to capture real time video images. And these video seq

吹牛者

growth-factor

https://doi.org/10.1007/978-3-322-83689-2tion. In the interview dialog, the system should ask about the subject that the user is interested in to obtain the user’s information efficiently. In this paper, we proposed the method to select the system’s utterance based on the user’s emotion to a focus detected from the user’s utterance. We pre

Seizure

Die Client/Server-Infrastruktur,shown high recognition performance. On the other hand, the speech recognition engine, Julius, has been widely used especially in Japan. Julius is also attracting attention since DNN-HMM is implemented in it. In this paper, we describe the design plan of interfaces that make Kaldi speech recognition

Injunction

https://doi.org/10.1007/978-3-322-86864-0ized by introducing adversarial learning to the conventional voice conversion. Adversarial learning is expected to enable us more natural voice conversion by using a discriminative model which classifies input speech to natural speech or converted speech in addition to a generative model. Experiment

遺傳學(xué)

cardiovascular

enflame

Technik der Client/Server-Architektur,arget is interrogation speech recorded during a police investigation. The fingerprint uses line spectral pairs (LSP) to measure the spectral envelope of the speech, and is coarsely quantized so that the fingerprint will not be altered by small degradation in the signal, but will be altered enough by