Looking into video frames on small displays
MULTIMEDIA '03 Proceedings of the eleventh ACM international conference on Multimedia
CODES/ISSS '10 Proceedings of the eighth IEEE/ACM/IFIP international conference on Hardware/software codesign and system synthesis
Adaptive encoding of zoomable video streams based on user access pattern
MMSys '11 Proceedings of the second annual ACM conference on Multimedia systems
Supporting region-of-interest cropping through constrained compression
ACM Transactions on Multimedia Computing, Communications, and Applications (TOMCCAP)
Towards understanding user tolerance to network latency in zoomable video streaming
MM '11 Proceedings of the 19th ACM international conference on Multimedia
Hi-index | 0.00 |
Modern mobile devices support multi-touch gestures that allow users to naturally zoom into and pan around Web pages, photos, and videos. When users zoom into a video, only part of the region in the video frames are displayed. Ideally, only the regions that the user is viewing are decoded, reducing the computation time (hence increasing the playback frame rate) and power consumption. We call this selective decoding. We have implemented a system consisting of an offline analyzer and a mobile video player that implements selective decoding in MPEG-4 Part 2 Simple Profile codec. The analyzer traces various dependency relationships among macroblocks of a given video and produces a meta-data file. The mobile video player supports zoom and pan gestures, and uses the meta-data to trace the macroblocks that are needed to decode the RoI. The player uses a modified decoding process to decode macroblocks selectively based on the trace. Our experiments show that selective decoding can improve playback frame rate by up to 193.3% and reduce energy consumption by up to 64.5%.