Eric Barrau

MPEG video transcoding to a fine-granular scalable format

The principle for elastic storage of compressed multimedia content has been introduced. It is based on the combination of a random-access storage medium and a bit rate scaling method. When the storage medium is full, the bit rate of the content that has been previously stored is reduced allowing additional content to be stored. For compressed video, the bit rate can be reduced thanks to well known bit rate transcoding techniques. However, these techniques suffer from complexity problems in the scope of elastic storage application and do not present good performance in terms of resulting picture quality because transcoding is applied several times on the same video. An alternative technique referred to as transcoding to a SNR scalable format is introduced in this paper. Based on the fine granular scalability (FGS) proposed in the MPEG-4 standard, it has been adapted to our transcoding problem. Compared to the bit rate transcoding approach, the proposed solution presents better resulting picture quality and does not suffer from complexity problems when bit rate reduction of multiple streams is required.

A scalable MPEG-2 bit-rate transcoder with graceful degradation

This paper describes a scalable MPEG-2 bit-rate transcoder that scales from low complexity, resulting in moderate picture quality, to high complexity with much better quality. The algorithm presents a very fine grain scalability in terms of complexity, making this solution especially suitable for software implementations, in order to benefit from the available resources to increase picture quality. Picture quality degrades gracefully when the complexity of the algorithm is decreased.

Low-resolution optimized 3D-subband scalable codec

Today, 3D subband (3DS) video coding schemes are close to current standard solutions in terms of coding efficiency while they add the scalability functionality through embedded bitstreams. Spatial scalability may be regarded as a key-feature brought by such codecs, enabling adaptation to varying terminal capabilities and display sizes. However, this functionality still suffers from a lack of coding efficiency when motion compensation is used. Though enabling motion compensation at the temporal filtering stage dramatically improves energy compaction, it generates a strong motion vector overhead and introduces a reconstruction drift when decoding at lower spatial resolutions. This is because the Discrete Wavelet Transform and the motion compensation (MC) are not commutative. Some authors solve this problem by transmitting the drift signal as side information, but this increases the bit-rate. In this paper we present a low-resolution optimized MC-3DS scalable codec with no drift, that does not generate any overhead. Its structure is fully compliant with any subband-tree entropy encoder and preserves all the other scalability functions (temporal and SNR). Tests and simulations show that the new scheme remains quite efficient when decoding at full resolution, while it outperforms the previous solution as far as spatial scalability is concerned, especially with high-activity sequences.