IP Library › Granted Patent US 12711968
Granted Patent B2
US 12711968 · App. 18/306,026 · Granted Aug 18, 2026

Multiple scrambled layers for audio watermarking

Inventors: Vladimir Kuznetsov (Ellicott City, MD); Sadhana Gupta (Columbia, MD); Wendell Lynch (East Lansing, MI)
Assignee: The Nielsen Company (US), LLC
G10L19/018G06T1/0028G06T1/0071
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12711968
App. No.
18/306,026
Granted
Aug 18, 2026
Kind
B2
Abstract

Apparatus, systems, articles of manufacture, and methods are disclosed for multiple scrambled layers for audio watermarking. An example system includes a processor to execute instructions to: divide the watermark into a plurality of watermark symbols; map the watermark symbols to a plurality of frequency bins according to a first distribution scheme to create a first watermark layer having first frequency bins and a second watermark layer having second frequency bins, subsets of adjacent ones of the plurality of frequency bins grouped into a corresponding plurality of frequency clumps; determine a sequence for shifting watermark symbols to cause different combinations of watermark symbols that include one watermark symbol from each watermark layer to share a first number of frequency bins, the first number to be less than a number of watermark symbols in the different combinations; and generate a second distribution scheme to map the watermark symbols.

Claims (48)

1 . A system comprising:

a processor; and

a non-transitory computer-readable storage medium comprising instructions that, when executed by the processor, cause the processor to:

map watermark symbols to first pluralities of frequency bins over a plurality of frequency clumps based on a first bin distribution corresponding to a first watermark layer;

map the watermark symbols to second pluralities of frequency bins over the plurality of frequency clumps based on a second bin distribution corresponding to a second watermark layer, the second bin distribution related to the first bin distribution based on a first sequence of shifts;

determine a second sequence of shifts different from the first sequence of shifts;

generate a third bin distribution corresponding to a third watermark layer based on the second sequence of shifts, the third bin distribution to map the watermark symbols to third pluralities of frequency bins over the plurality of frequency clumps;

communicate the third bin distribution to an encoder to enable the encoder to generate an encoded audio signal with the watermark symbols based on the third bin distribution; and

communicate the third bin distribution to a decoder to enable the decoder to extract the watermark symbols from the encoded audio signal.

2 . The system of claim 1 , wherein the instructions cause the processor to randomly determine the second sequence of shifts.

3 . The system of claim 1 , wherein the instructions cause the processor to:

determine if a first watermark symbol and a second watermark symbol share more than one frequency bin between the third watermark layer and at least one of the first watermark layer or the second watermark layer;

in response to a determination that the first watermark symbol and the second watermark symbol share more than one frequency bin, determine a third sequence of shifts different from the second sequence of shifts; and

generate the third bin distribution based on the third sequence of shifts.

4 . The system of claim 1 , wherein the first bin distribution maps a first watermark symbol to a first frequency bin of a first frequency clump and a second frequency bin of a second frequency clump in the first watermark layer, and maps a second watermark symbol to the first frequency bin of the first frequency clump and a third frequency bin of the second frequency clump in the second watermark layer, the third frequency bin different from the second frequency bin.

5 . The system of claim 4 , wherein the second bin distribution maps the first watermark symbol to a fourth frequency bin of the second frequency clump in the third watermark layer and maps the second watermark symbol to a fifth frequency bin of the second frequency clump in the third watermark layer, the fourth frequency bin different from the third frequency bin, the fifth frequency bin different from the fourth frequency bin.

6 . The system of claim 5 , wherein the second frequency bin and the third frequency bin are separated by a first distance and the fourth frequency bin and the fifth frequency bin are separated by a second distance different from the first distance.

7 . A non-transitory computer-readable storage medium comprising instructions that, when executed by one or more processors, cause the one or more processors to at least:

map watermark symbols to first pluralities of frequency bins over a plurality of frequency clumps based on a first bin distribution corresponding to a first watermark layer;

map the watermark symbols to second pluralities of frequency bins over the plurality of frequency clumps based on a second bin distribution corresponding to a second watermark layer, the second bin distribution related to the first bin distribution based on a first sequence of shifts;

determine a second sequence of shifts different from the first sequence of shifts;

generate a third bin distribution corresponding to a third watermark layer based on the second sequence of shifts, the third bin distribution to map the watermark symbols to third pluralities of frequency bins over the plurality of frequency clumps;

communicate the third bin distribution to an encoder to enable the encoder to generate an encoded audio signal with the watermark symbols based on the third bin distribution; and

communicate the third bin distribution to a decoder to enable the decoder to extract the watermark symbols from the encoded audio signal.

8 . The storage medium of claim 7 , wherein the instructions are to cause the one or more processors to randomly determine the second sequence of shifts.

9 . The storage medium of claim 7 , wherein the instructions are to cause the one or more processors to:

determine if a first watermark symbol and a second watermark symbol share more than one frequency bin between the third watermark layer and at least one of the first watermark layer or the second watermark layer;

in response to a determination that the first watermark symbol and the second watermark symbol share more than one frequency bin, determine a third sequence of shifts different from the second sequence of shifts; and

generate the third bin distribution based on the third sequence of shifts.

10 . The storage medium of claim 7 , wherein the instructions are to cause the one or more processors to embed a first watermark symbol and a second watermark symbol in an audio signal, the first watermark symbol corresponding to one of the first pluralities of frequency bins and the second watermark symbol corresponding to one of the third pluralities of frequency bins.

11 . The storage medium of claim 7 , wherein the first bin distribution maps a first watermark symbol to a first frequency bin of a first frequency clump and a second frequency bin of a second frequency clump in the first watermark layer, and maps a second watermark symbol to the first frequency bin of the first frequency clump and a third frequency bin of the second frequency clump in the second watermark layer, the third frequency bin different from the second frequency bin.

12 . The storage medium of claim 11 , wherein the second bin distribution maps the first watermark symbol to a fourth frequency bin of the second frequency clump in the third watermark layer and maps the second watermark symbol to a fifth frequency bin of the second frequency clump in the third watermark layer, the fourth frequency bin different from the third frequency bin, the fifth frequency bin different from the fourth frequency bin.

13 . The storage medium of claim 12 , wherein the second frequency bin and the third frequency bin are separated by a first distance and the fourth frequency bin and the fifth frequency bin are separated by a second distance different from the first distance.

14 . A method comprising:

mapping watermark symbols to first pluralities of frequency bins over a plurality of frequency clumps based on a first bin distribution corresponding to a first watermark layer;

mapping the watermark symbols to second pluralities of frequency bins over the plurality of frequency clumps based on a second bin distribution corresponding to a second watermark layer, the second bin distribution related to the first bin distribution based on a first sequence of shifts;

determining, by executing an instruction with a processor, a second sequence of shifts different from the first sequence of shifts;

generating, by executing an instruction with the processor, a third bin distribution corresponding to a third watermark layer based on the second sequence of shifts, the third bin distribution to map the watermark symbols to third pluralities of frequency bins over the plurality of frequency clumps;

communicating the third bin distribution to an encoder to enable the encoder to generate an encoded audio signal with the watermark symbols based on the third bin distribution; and

communicating the third bin distribution to a decoder to enable the decoder to extract the watermark symbols from the encoded audio signal.

15 . The method of claim 14 , wherein the second sequence of shifts is determined randomly.

16 . The method of claim 14 , further including:

determining if a first watermark symbol and a second watermark symbol share more than one frequency bin between the third watermark layer and at least one of the first watermark layer or the second watermark layer;

in response to determining that the first watermark symbol and the second watermark symbol share more than one frequency bin, determining a third sequence of shifts different from the second sequence of shifts; and

generating the third bin distribution based on the third sequence of shifts.

17 . The method of claim 14 , wherein the first bin distribution maps a first watermark symbol to a first frequency bin of a first frequency clump and a second frequency bin of a second frequency clump in the first watermark layer, and maps a second watermark symbol to the first frequency bin of the first frequency clump and a third frequency bin of the second frequency clump in the second watermark layer, the third frequency bin different from the second frequency bin.

18 . The method of claim 17 , wherein the second bin distribution maps the first watermark symbol to a fourth frequency bin of the second frequency clump in the third watermark layer and maps the second watermark symbol to a fifth frequency bin of the second frequency clump in the third watermark layer, the fourth frequency bin different from the third frequency bin, the fifth frequency bin different from the fourth frequency bin.

19 . The method of claim 18 , wherein the second frequency bin and the third frequency bin are separated by a first distance and the fourth frequency bin and the fifth frequency bin are separated by a second distance different from the first distance.