Adaptive Streaming of Wavelet-based Secure Scalable Video

The MPEG-21 Call for Contributions for a scalable video coding system (SVC) has been answered by a significant amount of wavelet-based proposals involving motion threading and temporal lifting, which represent the state of the art technology in wavelet-based scalable video coding. Multimedia security is a heavily researched topic at present, especially due to the financial interests of media content owners. Encryption and watermarking technologies for wavelet-based visual data formats in the context of digital rights management have been mainly focussed on JPEG 2000 (JPSEC) and still image coding schemes in general (due to the former lack of wavelet-based video coding standards). The first aim of this project is to develop scalable encryption and watermarking schemes specifically tailored to the corresponding wavelet-based video coding systems. In particular, partial or selective encryption schemes providing bitstream compliance for stream adaptation without decoding on the one hand and reducing the computational load for encryption on the other hand will be developed. As an example we mention the possible combination of partially encrypting texture data (i.e. frame data) and scalable motion data in an optimal way. Scalable watermarking techniques are required as well which can offer protection across a wide range of quality/resolution layers. Another goal of the project, seemingly conflicting to the first one, is to stream the secured scalable videos and adapt them to downstream client device capabilities, user preferences, or network conditions, if and where required. That is, universal scalability (temporal, spatial, and quality dimensions) will be considered. Adaptation can happen on a (trusted) server or on mid-network (possibly untrusted) nodes and will, in most cases, be performed as truncation of, or random access into, the scalable video bitstreams. The challenge in this project is therefore to perform these adaptation operations without compromising the streams` compliance and security properties (and thereby resolving the observed conflict with respect to satisfying both requirements).

The MPEG-21 Call for Contributions for a scalable video coding system (SVC) has been answered by a significant amount of wavelet-based proposals involving motion threading and temporal lifting, which represent the state of the art technology in wavelet-based scalable video coding. Multimedia security is a heavily researched topic at present, especially due to the financial interests of media content owners. Encryption and watermarking technologies for wavelet-based visual data formats in the context of digital rights management have been mainly focussed on JPEG 2000 (JPSEC) and still image coding schemes in general (due to the former lack of wavelet-based video coding standards). The first aim of this project is to develop scalable encryption and watermarking schemes specifically tailored to the corresponding wavelet-based video coding systems. In particular, partial or selective encryption schemes providing bitstream compliance for stream adaptation without decoding on the one hand and reducing the computational load for encryption on the other hand will be developed. As an example we mention the possible combination of partially encrypting texture data (i.e. frame data) and scalable motion data in an optimal way. Scalable watermarking techniques are required as well which can offer protection across a wide range of quality/resolution layers. Another goal of the project, seemingly conflicting to the first one, is to stream the secured scalable videos and adapt them to downstream client device capabilities, user preferences, or network conditions, if and where required. That is, universal scalability (temporal, spatial, and quality dimensions) will be considered. Adaptation can happen on a (trusted) server or on mid-network (possibly untrusted) nodes and will, in most cases, be performed as truncation of, or random access into, the scalable video bitstreams. The challenge in this project is therefore to perform these adaptation operations without compromising the streams` compliance and security properties (and thereby resolving the observed conflict with respect to satisfying both requirements).