PatentRef

Data version 20261027.2, as of 2026-10-27. Free beta.

US 10034028

Caption and/or metadata synchronization for replay of previously or simultaneously recorded live programs

As of 2026-10-27, the USPTO maintenance-fee record for US 10034028 has an entry dated 2026-08-31 recording that the patent expired for nonpayment of a maintenance fee, and no reinstatement after it (data version 20261027.2).

Status
expired_fee_nonpayment
Basis
fee_lapse:2026-08-31 term_20y (fee-lapse entry dated 2026-08-31, 20 years from the filing of a parent application on 2010-09-21)
Term end, computed from the record's dates
2030-09-21
Fee lapse
2026-08-31
Reinstatement on record
none
Reinstatement possible
unknown: 37 CFR 1.378: a petition to accept a delayed maintenance fee on the basis of unintentional delay has no fixed time limit; the fee-event record shows no reinstatement as of 2026-10-27
As of
2026-10-27
Data version
20261027.2

A petition under 37 CFR 1.378 has no fixed time limit; the record does not show whether one would be granted.

Family outside the US

Publications outside the US that Google Patents Public Datasets lists in the same patent family as US 10034028 (read 2026-10-15). Their legal status is not shown here.

OfficeNumberKindPublished
CA2774985A12011-03-31
CA2774985C2016-09-13
EP2481207A22012-08-01
EP2481207A42016-08-24
WO2011037951A22011-03-31
WO2011037951A32011-08-18

This patent's family on USPTO Global Dossier

Family data from Google Patents Public Datasets, used under CC BY 4.0.

Links

Record

{
 "patent_id": "10034028",
 "status": "expired_fee_nonpayment",
 "basis": [
  "fee_lapse:2026-08-31",
  "term_20y"
 ],
 "reinstatement_possible": "unknown",
 "reinstatement_basis": "37 CFR 1.378: a petition to accept a delayed maintenance fee on the basis of unintentional delay has no fixed time limit; the fee-event record shows no reinstatement as of 2026-10-27",
 "expiry_date_estimated": "2030-09-21",
 "lapse_date": "2026-08-31",
 "reinstated_date": null,
 "as_of": "2026-10-27",
 "data_version": "20261027.2",
 "catalog_data_version": "20261027.2",
 "permalink": "https://patentref.io/p/10034028?v=20261027.2",
 "next_steps": [
  {
   "kind": "record",
   "url": "https://patentref.io/p/10034028?v=20261027.2",
   "audience": "human"
  },
  {
   "kind": "patent",
   "url": "https://patentref.io/api/v1/patent/10034028/",
   "audience": "agent"
  },
  {
   "kind": "fee_events",
   "tool": "uspto_patent_center",
   "url": "https://patentcenter.uspto.gov/applications?patentNumber=10034028",
   "audience": "agent"
  }
 ]
}

Claims

The claims as the USPTO long-text file prints them (text data version 20261027.3), split into their parts by PatentRef's claim parser (structure version 2): each claim's text up to its first element, then one line for each element. Each claim and each element has a link of its own, made from the claim's number and, for an element, its place in the claim and its character offsets in the claim text, so a link stays the same as long as the data does. How the links are made.

Claim 1

Independent claim.

1. A method in a computer system for caption synchronization of media programs comprising

  1. Claim 1, element 1receiving at least one portion of a media stream and at least one portion of a caption data stream in the computer system,
  2. Claim 1, element 2wherein the at least one portion of the media stream and the at least one portion of the caption data stream correspond to an event; and
  3. Claim 1, element 3the at least one portion of the media stream comprises timing information and at least one audio component; and
  4. Claim 1, element 4utilizing at least one processing unit in the computer system to translate an audio signal into a sequence of word sounds in an audio stream;
  5. Claim 1, element 5divide the audio stream into a plurality of overlapping audio stream segments that each correspond to a respective time slice of the audio stream,
  6. Claim 1, element 6wherein each audio stream segment overlaps with a temporally adjacent audio stream segment such that temporally adjacent audio stream segments share a portion of the audio stream;
  7. Claim 1, element 7translate the audio stream into a string of words using a combination of a base language model and an event specific language model;
  8. Claim 1, element 8produce an audio-processed transcript by executing an indexing engine on the string of words,
  9. Claim 1, element 9wherein the audio processed transcript includes a text stream of at least a plurality of sequences of words and timing information;
  10. Claim 1, element 10align the audio-processed transcript with the at least one portion of the caption data stream; and
  11. Claim 1, element 11generate a time synchronized caption data stream by applying the timing information from the audio-processed transcript to the at least one portion of the caption data stream based on the aligning.

Claim 2

Depends on claim 1.

2. The method of claim 1 , wherein

  1. Claim 2, element 1the at least one portion of the caption data stream includes at least one metadata annotation and the time synchronized caption data stream includes text, and the method comprises searching at least one of text of the time synchronized caption data stream and the at least one metadata annotation.

Claim 3

Depends on claim 2.

3. The method of claim 2 , wherein

  1. Claim 3, element 1the event specific language model is based on a statistical representation of caption information provided in the at least one portion of the caption data stream.

Claim 4

Depends on claim 1.

4. The method of claim 1 , comprising

  1. Claim 4, element 1providing at least one search engine with access to the time synchronized caption data stream.

Claim 5

Independent claim.

5. A method comprising

  1. Claim 5, element 1translating, using an acoustic model engine of a first processing unit, a portion of an audio component of a media stream into at least one sequence of logical word sounds;
  2. Claim 5, element 2dividing the at least one sequence of logical word sounds into a plurality of overlapping audio segments,
  3. Claim 5, element 3wherein each audio segment overlaps with a temporally adjacent audio segment such that temporally adjacent audio segments share a portion of the sequence of logical word sounds;
  4. Claim 5, element 4converting, using a language model engine of a second processing unit, the at least one sequence of logical word sounds into a sequence of textual words;
  5. Claim 5, element 5wherein the language model engine is tuned based on a caption stream containing one or more portions of caption information; and
  6. Claim 5, element 6producing, using a third processing unit, a time based text by associating the at least one sequence of textual words with a time at which the portion of the audio component arises in the media stream.

Claim 6

Depends on claim 5.

6. The method of claim 5 , wherein

  1. Claim 6, element 1the caption stream is associated with the media stream.

Claim 7

Depends on claim 6.

7. The method of claim 6 comprising

  1. Claim 7, element 1providing at least one search engine with access to the time based text.

Claim 8

Depends on claim 7.

8. The method of claim 7 , wherein

  1. Claim 8, element 1at least two of the first processing unit, the second processing unit, and the third processing unit are executed by a single computing device.

Claim 9

Depends on claim 5.

9. The method of claim 5 , wherein

  1. Claim 9, element 1the language model engine includes a combination of a base language model and an event specific language model.

Claim 10

Depends on claim 5.

10. The method of claim 5 comprising,

  1. Claim 10, element 1in the acoustic model engine, recognizing at least one word from acoustic sounds based on an acoustic model based on a sampled recording.

Claim 11

Depends on claim 10.

11. The method of claim 10 , wherein

  1. Claim 11, element 1the at least one sample recording is based on an event type associated with the media stream.

Claim 12

Depends on claim 11.

12. The method of claim 11 , wherein

  1. Claim 12, element 1the language model engine includes a combination of a base language and an event specific language model.

Claim 13

Depends on claim 12.

13. The method of claim 12 , wherein

  1. Claim 13, element 1the event specific language model is selected based on caption information provided in the caption stream.

Claim 14

Depends on claim 12.

14. The method of claim 12 , wherein

  1. Claim 14, element 1the event specific language model is based on a statistical representation of the caption information provided in the caption stream.

Claim 15

Depends on claim 5.

15. The method of claim 5 comprising:

  1. Claim 15, element 1aligning the time indexed text with at least one of the one or more portions of caption information; and
  2. Claim 15, element 2applying timing information in the time indexed text to at least one of the one or more portions of the caption information aligned to the time indexed text.

Claim 16

Depends on claim 15.

16. The method of claim 15 , wherein

  1. Claim 16, element 1the aligning of the time indexed text with at least one of the one or more portions of caption information comprises:
  2. Claim 16, element 2dividing the one or more portions of caption information into a plurality of discrete caption data segments;
  3. Claim 16, element 3comparing at least one of a plurality of text segments based on the plurality of overlapping audio segments to a least one of the plurality of discrete caption data segments;
  4. Claim 16, element 4associating the at least one of the plurality of text segments with the at least one of the plurality of discrete caption data segments if the comparison indicates a correlation exists between the at least one of the plurality of text segments and the at least one of the plurality of discrete caption data segments.

Claim 17

Independent claim.

17. A method for identifying content related to a media stream based on a search of a database containing time synchronized caption data streams comprising

  1. Claim 17, element 1inputting a text based search query into a search engine,
  2. Claim 17, element 2wherein the search engine is configured to identify one or more segments of a time synchronized caption data stream responsive to the query,
  3. Claim 17, element 3wherein the time synchronized caption data stream is generated pursuant to the methodology of claim 1 ;
  4. Claim 17, element 4in response to the identification of the one or more responsive segments of the time synchronized caption data stream, extracting at least one time based location in the media stream which corresponds to the query;
  5. Claim 17, element 5providing access to the at least one time based location in the media stream; and
  6. Claim 17, element 6upon accessing the at least one time based location in the media stream, presenting the portion of the media stream, at the identified time based location, on a presentation device.

Claim 18

Depends on claim 1.

18. The method of claim 1 , wherein

  1. Claim 18, element 1the aligning step further comprises:
  2. Claim 18, element 2dividing the at least one portion of the caption data stream into a plurality of discrete caption data segments;
  3. Claim 18, element 3selecting a first group of two or more audio stream segments;
  4. Claim 18, element 4comparing the first group of audio stream segments to a first caption data segment to determine if a correlation exists between the first caption data segment and at least one of the two or more audio stream segments in the first group of two or more audio stream segments;
  5. Claim 18, element 5if the correlation does not exist, assessing whether at least one of the two or more audio stream segments in the first group of two or more audio stream segments has an association with a successive caption data segment; and
  6. Claim 18, element 6if the association does not exist, repeating the assessing step until an association is found between at least one of the two or more audio stream segments and at least one of the plurality of discrete caption data segments.