Patent Application
20130070960
The invention relates to a method and to a system for providing a watermarked decoded audio or video signal derived from a watermarked audio or video signal that was encoded and decoded.
Various audio signal watermarking (WM) systems are known, which are also applied to digitally encoded audio signals.
However, in current audio signal watermarking systems the watermark signal and its watermark payload often do not survive very low bit rate audio coding, in particular parametric encoding. This means that, following compression of a watermarked audio signal by such a coder, the watermark payload cannot be retrieved any more at receiver or decoder side.
This causes a problem because parametric audio coding (like AAC+, mp3PRO, unified speech and audio coder) is going to be used more and more in the future.
In one usage scenario, a dedicated company performs post production and authoring of pre-release DVDs for a studio. The studio requires that the content is watermarked, so that a leak can be tracked if content is leaked to the internet. Specific-language subtitles for the movie are to be generated by a subcontractor in a country where that specific language is spoken. Therefore the original-language audio content is sent to that subcontractor. Since using high bandwidth is expensive, the audio signal is compressed strongly by a codec using parametric coding. As a consequence, the watermark can no more be retrieved from the content the subcontractor has received. The dedicated company can comply with its contract with the studio only by sending high-quality or nearly high-quality (i.e. weak compression) audio content to the subcontractor, which is more expensive and takes more time.
A problem to be solved by the invention is to provide watermarked and—in particular low bit rate—encoded and decoded audio signals from which the watermark signal and the watermark payload can be detected at decoder side. This problem is solved by the methods disclosed in claims 1 to 3. A corresponding system is disclosed in claim 4.
According to the invention, a watermarked uncompressed audio signal bitstream is—in particular low bit rate—encoded, but the watermarking signal is retrieved (and may be removed in the encoded audio signal), and the decoded watermark signal payload only is transmitted or transferred together with the encoded audio signal bitstream. At decoder side, the received audio signal bitstream is de-compressed and watermarked again using the transmitted watermark payload.
The invention combines the advantages of parametric or other low bit rate encoding like small audio signal file size with the advantages of watermarking like traceability of content.
In principle, the inventive method is suited for providing a watermarked decoded audio or video signal derived from a watermarked audio or video signal that was low bit rate encoded and decoded, said method including the steps:
encoding said watermarked audio or video signal and retrieving the watermark signal so as to provide an encoded audio or video signal;
decoding said retrieved watermark signal so as to provide the included watermark payload data;
encoding said watermark payload data so as to provide a related watermark signal;
decoding said encoded audio or video signal, thereby embedding said watermark signal, so as to provide a decoded watermarked audio or video signal.
In principle the inventive system is suited for providing a watermarked decoded audio or video signal derived from a watermarked audio or video signal that was low bit rate encoded and decoded, said system including:
means being adapted for encoding said watermarked audio or video signal and retrieving the watermark signal so as to provide an encoded audio or video signal;
means being adapted for decoding said retrieved watermark signal so as to provide the included watermark payload data;
means being adapted for encoding said watermark payload data so as to provide a related watermark signal;
means being adapted for decoding said encoded audio or video signal, thereby embedding said watermark signal, so as to provide a decoded watermarked audio or video signal.
Advantageous additional embodiments of the invention are disclosed in the respective dependent claims.
Exemplary embodiments of the invention are described with reference to:
In
Preferably, the audio signal encoding and decoding is a low bit rate encoding and decoding. The term “low bit rate” means a bit rate which is so low that an additional watermark embedded in the code for the encoded audio or video signal, following decoding of the audio or video signal, can no more be detected or evaluated usefully, i.e. an encoded audio signal data rate below a range of e.g. 24 kBit/s to 6 kBit/s. If any remaining watermark signal parts could disturb the watermarking at receiving side, the watermark signal is removed when encoding the audio signal. As an alternative, the audio signal encoding and decoding is not a low bit rate encoding/decoding in the above meaning. In such case the watermark signal must be removed when encoding the audio signal.
Advantageously, the data rate of the encoded audio signal will not increase significantly from the embedding of the WM payload data because the data rate of the WM payload data WMP is much smaller than the data rate of the audio coding parameters representing the encoded audio signal EAS. The inventive processing is independent from the specific parametric encoder/decoder used because it is carried out in the uncompressed domain.
The uncompressed audio signal is always watermarked. Since it is to be decompressed before it becomes useful (e.g. for listening to it), there is kind of protection in the non-watermarked compressed domain.
The inventive processing operates well under the condition that the communication chain audio encoder/WM detector 11, transmission channel and audio decoder/WM embedder 13 can be controlled completely. In case the transmission channel or line cannot be controlled and a standard audio codec is used, an adversary could read the bitstream and use a standard audio decoder without WM embedder to create a non-watermarked version of the audio signal. Therefore the audio encoder/WM detector can also include a known encryption module, so that the bitstream EAS and possibly the watermark payload WMP is encrypted. The audio decoder/WM embedder 13 includes a corresponding decryption module, so that the encrypted bit stream is decrypted, decoded and watermarked. The watermark is embedded jointly with the deciphering and decompression, such that the watermark embedding during the decoding is rendered mandatory because it is linked to the deciphering. This represents a conversion from parametric data (parametric compressed data, watermark) to an integrated multimedia signal (watermarked decoded audio) in a secure way.
The audio encoding/decoding is e.g. an AAC+, Unified Speech and Audio coder, CELP, MPEG4 Parametric Audio, SBR Parametric Stereo (PS), or mp3PRO encoding/decoding.
Instead of an audio signal, the invention can relate to a video signal, wherein the low bit rate audio encoding/decoding is replaced by a low bit rate video encoding/decoding.
The audio or video signal can be encoded/decoded frame-wise, wherein the watermark signal sections are assigned to these frames.
| Number | Date | Country | Kind |
|---|---|---|---|
| 10305579.4 | Jun 2010 | EP | regional |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/EP2011/058037 | 5/18/2011 | WO | 00 | 11/30/2012 |
