News

Hyunsuk Ko successfully passed his Qualifying exam on Jan 28th. The title is “A ParaBoost Stereoscopic Image Quality Assessment (PBSIQA) System”. His Qual Exam Committee includes: Jay Kuo (Chair), Antonio Ortega, Keith Jenkins, Sandy Sawchuk and Aiichiro Nakano (Outside Member).

The design of a novel ParaBoost (parallel-boosting) stereoscopic image quality assessment (PBSIAQ) system based on machine learning is presented in his work. The PBSIQA system consists of two stages. In the first stage, they classify multiple distortions into multiple types, and develop a set of individual quality scorers targeting at specific distortion types and working in parallel in the first stage. In the second stage, to obtain the best overall performance against the whole database, they adopt the parallel boosting idea from machine learning. Extensive experimental results are provided to demonstrate that the superb performance of the proposed PBSIQA scorer, which outperforms existing stereo image quality assessment (SIQA) indices by a significant margin.

Multimedia Communication Lab held the first internal screening exam – programming from 1/7/2014 to 1/9/2014. The exam includes three parts: written examination, oral examination and practical computer-based programming with languages limited in C/C++/Java/Python. As a complement of the official screening exam, this internal exam aims to enhance the programming ability of Ph.D candidates. Each test taker is supposed to accomplish a project within three days, including implementing a state-of-the-art algorithm on selected topic and submitting final report with detailed discussion about the project.

Three group members took the exam this year and congratulations to those who passed.

The result of Golden Eye Test has come out shortly after the subjective tests for MCL-3D and MCL-V database finished, and six students won their awards according to their accurate observations on 3D images and video clips, respectively.
MCL-3D database has 9 image sets, each set has 77 images. In addition, there are 12 video sets in MCL-V database, and each set contains 9 video clips. Since we have collected 30 opinions scores for each set, there are 20790 opinion scores for MCL-3D database and 3600 for MCL-V database in total. Both databases will be released to public in short term.
For our future research, we will develop a stereoscopic image quality assessment algorithm on MCL-3D database for Samsung and an effective video quality assessment metric on MCL-V database for Netflix.

The Media Communications Lab had a wonderful Thanksgiving lunch together as a big family. We shared a very happy time.

Wireless physical layer secrecy has attracted much attention in recent years due to the broadcast nature of the wireless medium and its inherent vulnerability to eavesdropping. This book introduces various signal processing approaches to enhance physical layer secrecy in multi-antenna wireless systems. It focuses specifically on the signal processing aspects, including beamforming and precoding techniques for data transmission and discriminatory training schemes for channel estimation. [expand title=”Read More…” swaptitle=”See Less…”] The content can be roughly divided into three parts: (i) data transmission, (ii) channel estimation and (iii) advanced applications. Even though many works exist in the literature on these topics, the approaches and perspectives taken were largely diverse. The discussions will cover cases with collocated and distributed antennas, i.e., relays. This book will also review recent works that apply these signal processing approaches to more advanced wireless systems, such as OFDM systems, multicell systems, cognitive radio, multihop networks etc. This book will allow readers to gain basic understanding of works on physical layer secrecy, knowledge of how signal processing techniques can be applied to this area, and the application of these techniques in advanced wireless applications. The authors hope to provide a more organized and systematic view of these designs and to lay a solid foundation for future work in these areas. By presenting the work from a signal processing perspective, this book should trigger more research interest from the signal processing community and further advance the field of physical layer secrecy along the described directions.[/expand]

More information about this book:
Springer
USC Library