[행사/세미나] 11/30 실감미디어공학과 세미나 <Next-level Challenges to Raising the Bar on Video Streaming Efficiency and Quality>
- 실감미디어공학과
- 조회수1273
- 2023-12-05
안녕하세요
실감미디어공학과 사무실입니다.
실감미디어공학과 세미나가 아래와 같이 개최되오니 많은 참여를 부탁드립니다.
* 현재 본 세미나 수업을 수강하고 있지 않은 분들도 온/오프라인 세미나 참석이 가능합니다:)
▶세미나 접속 ZOOM 링크: 실감미디어공학과 세미나
https://us02web.zoom.us/j/5913891581?pwd=OHBSekRDdHhZa3dsVXYzL245bmJPZz09
회의 ID: 591 389 1581
암호: 908353
Colloquium on
Immersive Media Engineering
Dec. 8, 2023(Fri.) 16:30pm –18:00pm
International Hall, Room 90104
Dr. Rahul Vanam (Senior Research Scientist/Prime Video. Amazon Inc.)
Abstract:
In the last decade there has been a significant shift in viewership from cable TV to streaming services globally. Viewers stream videos over different network bandwidths, use different playback devices from smart phones to TVs having different decoding capabilities, making it challenging to lower bit-rate and improve visual quality across the board. Most devices have been deployed with decoders from older video coding standards like AVC and HEVC, and hence there is a need to improve the compression performance of existing encoders. For example, on film grain content, such encoders operate at the peak bit-rate and do not yield much compression gain. While the compression of grainy content can be improved using film grain synthesis (FGS), there are encoder challenges related to preserving signal and grain quality. The first part of the talk will cover such encoder related challenges. Video quality metrics (VQMs) are being used to design adaptive bitrate ladders for streaming, and the use of an appropriate VQM can help design optimal rate-quality ladders. Existing VQMs have limitations and are still not yet able to guide encoder subjective quality tuning process and fully automate content-adaptive encoder parameter selection. The limitations of VQMs and areas for improvement will be covered in the second part of the talk.
Biography:
Dr. Rahul Vanam is a Senior Research Scientist at Amazon Prime Video, Seattle, USA since 2020, where he conducts research related to video compression and video quality metrics. From 2010 to 2019, he was a Senior Staff Engineer at InterDigital Communications, USA, where he conducted research in perceptual video coding and the Versatile Video Coding (VVC) standard. From 2019 to 2020, he was a Principal Engineer at Brightcove, USA, where he developed perceptual video quality models and metrics. He has served as an Associate Editor for the IEEE Transactions on Image Processing from 2015 to 2018. He received the Ph.D. degree in electrical engineering from the University of Washington, Seattle, USA.