Duc V. Nguyen

Buffer-based bitrate adaptation for adaptive HTTP streaming

Live and on-demand video streaming has recently contributed a significant part of the Internet traffic. In order to guarantee the best possible user experience, multimedia service is required to support large fluctuations in different network conditions. One of the cost effective means for multimedia delivery nowadays is HTTP streaming. The key challenge of this technology is how to change adaptively video bitrates according to network variation and client buffer while bestowing upon users the best possible video quality. In this paper, we propose a novel rate-adaptation method on client side, which allows clients to estimate buffer level in the future, and thus achieving a balance between the need for high video quality and buffer stability. The experiments show that our method performs well under varying network conditions.

Quality-delay tradeoff optimization in multi-bitrate adaptive streaming

This paper studies the delay-quality tradeoff of variable bitrate (VBR) videos in multi-bitrate adaptive streaming. A novel method is presented that provides an adapted video with the best possible quality while still meeting the constraints of bandwidth and delay.

Request adaptation for adaptive streaming over HTTP/2

In this paper, we propose a client-based adaptation method for Adaptive Streaming over HTTP/2 leveraging the server push feature. A cost function that takes into account both the number of pushed segments and the playout buffer level is defined. Experiment results show that the proposed method can improve the tradeoff between the number of requests and buffer stability compared to existing solutions.

Future buffer based adaptation for VBR video streaming over HTTP

HTTP streaming has become a cost effective means for video delivery nowadays. To enable adaptivity to networks and terminals, a provider should generate multiple representations of an original video as well as the related signaling metadata. So far, most previous studies have just focused on the case of CBR (constant bit rate) video. In this paper, we propose a novel adaptation method for VBR (variable bitrate) video streaming. Based on a trellis representation to estimate future buffer levels, the proposed method can provide smooth video quality while avoiding buffer underflows. The experimental results show that our approach can perform effectively under drastic changes of both connection throughput and video bitrate.

Design and implementation of an e-Health system for depression detection

We present the design and implementation of a cost-effective e-Health system for automatic depression detection. The system is based on a client-server architecture, where clients are popular mobile devices. For practical deployment, various factors that affect the accuracy and speed of depression detection are discussed and evaluated with extensive experiments.

An Evaluation of Screen Content Casting over Mobile and Wireless Networks

Thanks to the availability of many smart devices, displays, and broadband connections, screen casting/sharing has become an important functionality for user devices. Meanwhile, a new video standard called screen content coding (SCC) is going to be issued to support this trend. This paper presents a standard-compliant implementation and evaluation of SCC streaming to mobile devices. The study is targeted at a feasible design and settings for customer devices, supporting both content-generating side and content-consuming side. The evaluation results help to answer how a sending device should generate screen content videos, and how a receiving device could be implemented and adjusted in time-varying environments of mobile networks and wireless home networks. To the best of our knowledge, this is the first evaluation study for SCC casting.

An adaptive streaming method of 360 videos over HTTP/2 protocol

Nowadays, 360-degree video has become an important component of virtual reality (VR) technology. Unfortunately, this state-of-the-art content type requires a huge bandwidth from a server to a client. In order to solve the bandwidth problem in streaming VR videos over HTTP/2, a dynamic adaptation method is crucial. In this paper, we propose a novel adaptive streaming method based on tiled streaming. By using H.265 standard, a video at the server is divided into spatial tiles, each of which is subdivided into multiple temporal segments. In order to support adaptive streaming method from client, each tile is also encoded into different versions. The priority of tiles is defined based on the users viewport. Then, the priority feature of HTTP/2 is used to request the server to push the tiles of higher priority first. This ensures that the visible tiles will arrive at the client and be decoded first. Besides, in the case of bandwidth fluctuation, we use the stream termination feature to cancel pushing the tiles that would arrive at the client after their playout deadlines.

A QoS-adaptive framework for screen sharing over Internet

In this paper, we present a QoS-adaptive framework for screen sharing over heterogeneous networks. Basic requirements and important issues of a screen sharing system are described. For determining optimal video coding parameters, adaptation trajectory is used as the hints in making decisions. Further, experiments are carried out to investigate the trade-off between video bitrate and the buffering delay in time-varying networks.

Quality-energy aware adaptation for mobile streaming clients

Energy saving is an important requirement for any mobile devices. Meanwhile, streaming is one of the most popular services via mobile devices now. In this paper, we study the dependence of consumed power on adapted video bitrates. A novel adaptation method is then presented to balance the requirements for high video quality and low power consumption.

Bandwidth management for multiple HTTP streaming clients

In this paper, we propose a solution for sharing bandwidth among multiple HTTP streaming clients running on different devices in a vehicular network. Our solution consists of two main aspects: a manager that can determine bandwidth allocated for each client and a client-based throughput control mechanism that can regulate throughput of each client.