FOXTEL Management Pty Ltd v TiVo, Inc
[2014] APO 7
•17 February 2014
IP AUSTRALIA
AUSTRALIAN PATENT OFFICE
FOXTEL Management Pty Ltd v TiVo, Inc.[2014] APO 7
Patent Application: 2006344734
Title:Method and Apparatus for Creating and Viewing Customized Multimedia Segments
Patent Applicant: TiVo, Inc.
Opponent: FOXTEL Management Pty Ltd
Delegate: Matt Kraefft
Decision Date: 17 February 2014
Hearing Date: 12 December 2013
Catchwords: PATENTS – section 59 – opposition to grant of a patent – novelty – opposition successful on the grounds of lack of inventive step, usefulness, clarity and fair basis – costs awarded against the applicant.
Representation: Patent applicant: Did not appear at the hearing
Opponent:Mr Guy Tucker and Mr Robert Miller, patent attorneys, Spruson & Ferguson, Sydney.
IP AUSTRALIA
AUSTRALIAN PATENT OFFICE
Patent Application: 2006344734
Title:Method and Apparatus for Creating and Viewing Customized Multimedia Segments
Patent Applicant: TiVo, Inc.
Date of Decision: 17 February 2014
DECISION
Claims 1-18, 23, 25 and 28 do not have an inventive step. Claims 1 and 3 are not useful. Claims 2 and 13 are not clear and claims 12, 21 and 22 are not fairly based.
The applicant is allowed two (2) months from the date of this decision in which to propose suitable amendments to overcome the above deficiencies.
REASONS FOR DECISION
Background
TiVo, Inc. (“the applicant”) filed patent application 2006344734 on 23 June 2006. The application claims priority from a US application filed on 22 June 2006. Application 2006344734 was advertised accepted on 24 March 2011.
FOXTEL Management Pty Ltd (“the opponent”) filed a notice of opposition on 23 June 2011. A statement of grounds and particulars in support of the opposition followed on 23 September 2011. The opponent subsequently amended this statement under Patent Regulation 5.16. This amendment was allowed on 16 September 2013.
The opponent completed its evidence-in-support, with several allowed extensions of time, on 23 August 2012. The applicant was also allowed several extensions of time to serve evidence-in-answer. In the end the applicant did not serve any evidence-in-answer. For the hearing, the applicant also chose not to file written submissions or attend the hearing.
The specification
The specification states the invention relates to recognizing in-band data in a multimedia content stream and locating points within the multimedia content stream at a multimedia device.
Digital video recorders (“DVRs”), or similar multimedia devices, record broadcasted television (“TV”) programs for later playback by changing electrical signals, digital or analogue, of the TV program into digital information and storing the digital information on a hard drive. The use of DVRs for recording TV programs has led to content providers searching for different means of presenting content to DVR users and enriching the experience of video or TV programs. The term DVRs includes set-top boxes (“STBs”) with a recording facility.
The specification states that a technique for recognizing distinct events within video programs is desirable. Content providers that desire to present additional content, such as marketing or enhanced textual content, within video programs can have particular icons, messages or interactive elements displayed to the user at particular parts of the video or TV program. A DVR must thus synchronize with the multimedia content stream to display the additional content data at the correct time.
The additional content may be in the form of in-band data. In-band data is used for signalling additional data over the same channel or band as the one used to send the audio and video data. Examples of in-band data include closed-caption data and Enhanced TV (“ETV”) signalling data. The former may present captions associated with the video content. The latter enables interactive applications.
The specification, as accepted, ends with 28 claims defining the claimed invention. There are two independent claims, not counting the omnibus claims. Those two claims are claims 1 and 3. Claims 1 and 3 are reproduced below.
- A method for creating and viewing customized multimedia streams, comprising:
receiving a first content stream at a first device, wherein the first content stream contains in-band data;
displaying the first content stream to a user;
receiving input from the user selecting one or more locations within the first content stream; and
in response to receiving the input from the user:determining a sequence of a plurality of hash values from in-band data corresponding to a user-selected location within the first content stream;
generating metadata associated with the user-selected location; and
storing the sequence of hash values and metadata as event identification data on a storage device.- A method for creating and viewing customized multimedia streams, comprising:
reading a first content stream from a removable storage medium at a first device, wherein the first content stream contains in-band data;
displaying the first content stream to a user;
receiving input from the user selecting one or more locations within the first content stream; anddetermining a sequence of a plurality of hash values from in-band data corresponding to a user-selected location within the first content stream;
generating metadata associated with the user-selected location; and
storing the sequence of hash values and associated metadata as event identification data on a storage device.Statement of Grounds and Particulars
As a consequence of the Intellectual Property Legislation Amendment (Raising the Bar) Act 2012, there are substantial changes to the Patents Act 1990. In the case of the present application the grounds of opposition are governed by the Patents Act in force before 15 April 2013.
The opponent’s grounds of opposition were that the nominated person is not entitled to the grant of a patent for the alleged invention, that the invention is not patentable because of a lack of manner of manufacture, novelty, inventive step and utility, that the specification does not fully describe the invention, and that the claims are not clear and fairly based. The opponent provided particulars against each of these grounds.
The grounds pressed at the hearing were all of the above except for the entitlement ground.
Evidence-In-Support
The opponent served evidence-in-support from Dr Jason Pieloor. At the time of his declaration, Dr Pieloor had about 16 years of experience as an engineer in product development and technology solutions for embedded platforms, primarily for subscription and pay-per-view television delivery.
In his evidence, Dr Pieloor has set out the state of the art related to multimedia content and its manipulation, particularly to TV or video broadcasting. He has further presented evidence related to the grounds of opposition pressed by the opponent at the hearing except for the manner of manufacture ground.
Claims construction
Claim 1 of the application defines receiving input from the user selecting one or more locations within the content stream. Furthermore, in response to receiving the input from the user, a further three steps of the method are performed. These steps consist of determining a sequence of a plurality of hash values from in-band data corresponding to a user-selected location, generating metadata associated with the user-selected location, and storing the sequence of hash values and the metadata. It is noteworthy that these steps, or the data arising therefrom, are all associated with a user-selected location. That is, claim 1 defines these steps being performed over a subset only of the content stream, that being only the user-selected location(s). There is no action associated with unselected locations. The claim equates to determining sequences of hash values only of bookmarks created in the content stream by users and to the generation of metadata associated only with those bookmarks.
Claim 3 does not include the expression “in response to receiving the input from the user”. On the other hand, claim 3 defines the same relationship of sequences of hash values and metadata with user-selected locations as claim 1. Claim 3 also defines a step of receiving input from the user selecting one or more locations. Consequently I conclude the feature of responsiveness to user input is inherent in claim 3. Similarly, claim 3 equates to determining sequences of hash values only of bookmarks and to the generation of metadata associated only with those bookmarks.
The above discussion would appear to influence the scope of the term “user” in the claims of the application. The opponent submitted there should be no limitation on the term and that the term could refer to a multimedia data content provider or enhancer as much as to an end user such as a home viewer for example. On the other hand, the above interpretation of claims 1 and 3 suggests a content provider or enhancer is not envisaged as the “user” in claims 1 and 3. A content provider as the “user” appears to be incongruent with the defined functions of determining sequences of hash values only of bookmarks and to the generation of metadata associated only with those bookmarks. Furthermore the claims and the description refer to customized multimedia streams. In context I conclude the term “user” in the claims refers to an end user such as a home viewer for example.
For completeness, I will also address the meaning of the expression “in-band data”. This is because the meanings ascribed by the applicant and by the opponent’s Dr Pieloor appear to be different.
At [0036] of the specification, the applicant defines in-band data as signalling data sent over the same channel or “band” as the one used to send audio and/or video data. Dr Pieloor, at [19] of his declaration, defines in-band data as data that is situational in time to the content. By this definition, audio and video data would be examples of in-band data as they primarily define the content stream. Subtitles would also be in-band data as they need to be synchronised with the audio and video content. According to Dr Pieloor, by contrast, out-of-band data is data that is not time-situational with the content. For example, metadata about the content, such as titles or lists of actors, would be out-of-band data since it does not share a common time-base with the content itself.
It is clear that some data types would be grouped in the same categories regardless of which definition is used. For example, subtitles or closed-caption data would be regarded as in-band data under both definitions. Nonetheless the appropriate definition should be resolved.
It appears that Dr Pieloor’s definition is somewhat artificial. That definition creates a dependency on whether the data itself is time-based or not. On the other hand, the applicant’s definition is more aligned with transmission of the multimedia data, the subject of the presently claimed invention. In the context of the present application, I would regard in-band data as the non-video or non-audio data flowing over the same path, channel or “band” as the video and/or audio data. Conversely, an out-of-band implementation would separate the video/audio data flow from the non-video and non-audio data flow.
The specification may also be set up as a dictionary in which the meaning of words is defined. In such cases the dictionary meaning or special meaning from the specification would prevail over any plain meaning from the art (British Thomson-Houston Company Ltd v Corona Lamp Works Ltd, [1922] 39 RPC 49 at 67). I think the specification at [0036] adequately sets up a dictionary for “in-band data”, amongst other terms.
For the above reasons, I accept the applicant’s definition for “in-band data”.
Patentable Invention
Section 18 of the Patents Act relates to patentable inventions. Relevant parts of subsection (1) appear below.
(1)Subject to subsection (2), an invention is a patentable invention for the purposes of a standard patent if the invention, so far as claimed in any claim:
(a) is a manner of manufacture within the meaning of section 6 of the Statute of Monopolies; and
(b) when compared with the prior art base as it existed before the priority date of that claim:
(i)is novel; and
(ii)involves an inventive step; and
(c) is useful; and …….
Novelty
In referring to several earlier decisions in respect to novelty, the Bristol-Myers Squibb Company v FH Faulding & Co Ltd decision, (2000) 46 IPR 553 at 576, states that what the authorities contemplate is that a prior publication, if it is to destroy novelty, must give a direction or make a recommendation or suggestion which will result, if the skilled reader follows it, in the claimed invention.
There may be circumstances where a document can deprive a claimed invention of novelty even though it does not expressly make that complete disclosure of all the integers of the claimed invention. One example is where the skilled reader would understand the disclosure of the prior publication to include a missing integer. Another is where the document contains a direction to use a process that inevitably or inexorably results in something within the claim (Danisco A/S v Novozymes A/S (No.2), [2011] FCA 282 at [248]).
D1 – US 2004/0040042 - Feinleib.
This document discloses a method for synchronizing enhancing content with primary content such as a video program. Closed captioning data for a program is broadcast in line 21 of the Vertical Blanking Interval (“VBI”), the VBI being the first 21 lines of a television frame. The closed captioning script is parsed to identify one or more key phrases. The parser may return a set of unique phrases with each phrase having the same number of words, or a unique data character string with each character string having the same length of characters.
An enhancing content producer decides at what points in a video program to introduce enhancing content. The enhancing content may be a hyperlink to a resource on the Internet, a manipulation of the video window with concurrent text or graphical data, or the launch of an application. When the placement of the enhancing content is determined in the video program, an authoring computer is used to associate supplemental data (e.g. a URL or a file name), for activating the enhancing content, with specific key phrases or character strings of the closed captioning script that correspond to the desired points in the program. The authoring computer creates a key phrase data file that contains a listing of the key phrases or character strings and their association to the supplemental data.
The key phrase data file is delivered to viewer computing units (e.g. a client system, personal computer, or TV with set-top box). When the program is played, a viewer computing unit with the key phrase data file for that program begins to monitor the closed captioning script. The viewer computing unit has a parser that identifies key phrases by examining sequential groups of words or characters and comparing them with the key phrases in the key phrase data file. Upon identification of a particular key phrase, the parser retrieves the supplemental data associated in the key phrase data file with the particular key phrase. The parser executes an enhancement action according to the supplemental data to synchronize the enhancement action with scenes in the video program.
With reference to Dr Pieloor’s declaration at [137] – [143], the opponent submitted that D1 discloses all of the features of at least claims 1 and 3 of the application. Dr Pieloor states that D1 discloses a content enhancing producer obtaining a copy of closed-caption script which is loaded into an authoring computer. Dr Pieloor equates the authoring computer to the first device of claim 1 of the application. Furthermore Dr Pieloor suggests the content enhancing producer may be anyone who wishes to enhance the content stream, including somebody independent of the producer of the primary content. Additionally the producer or enhancer can identify spots in a content stream by clicking a mouse button at the particular place as the script is running. Clearly Dr Pieloor equates the content producer or enhancer as a user.
It may be said there is selectivity in D1 of particular locations in a content stream by a content producer or enhancer. On the other hand, the General Tire & Rubber Co. v Firestone Tyre & Rubber Co. Limited decision, [1972] RPC 457 at 486, requires the following:
"To anticipate the patentee's claim the prior publication must contain clear and unmistakeable directions to do what the patentee claims to have invented…….. A signpost, however clear, upon the road to the patentee's invention will not suffice."
D1 does not clearly disclose the bookmarking of a content stream by an end user as defined in the context of the present claims.
Moreover, D1 identifies key phrases or character strings of closed captioning script at the selected locations rather than determining sequences of hash values. Dr Pieloor noted that D1 states at [0076] that the authoring computer might compute hashes of the key phrases if desired to reduce the size of the data file and improve matching speed at the client. However there are no clear and unmistakeable directions to determine sequences of hash values at selected locations as claimed in the present application, only that hashes might be computed of each word group for comparison to stored hash values.
I conclude the claimed invention is novel over D1.
D2 – US 2005/0226601 – Cohen et al.
This document discloses a method for synchronizing special effects to a media presentation. A media recording device may include one or more recording/reading units and be linked to a media source. Similarly a media presentation device may be linked to a media source. One or both of the media recording device or media presentation device may further be linked to an effects signal generator interface. The latter links to effects generators. The generators may include a scent generator, a smoke machine, a horn, bell, alarm, bubble blower, colour or lights generator, music generator, or other generator of effects for enhancing or supplementing a listening or viewing experience of a media presentation.
A processor in one or both of the above devices may execute algorithms for generating an identifying value from a segment or multiple segments of data in a data stream. The algorithms may be a cyclic redundancy check (“CRC”) algorithm or a hash algorithm. Furthermore the algorithms may be executed on close caption signals in a data stream.
The media recording device may include a time stamper for associating a time value with the segments in a data stream. For example the time stamper may assign a time stamp or time point to a position of one or more close caption signals in a presentation. The above-identified values may be loaded into a first table such that the value for a particular portion of a closed caption line is associated with the time stamp entry in the table that matches when the closed caption line appears in the recording. A second table may be created by a content provider to associate time stamps with effects signals so as to generate a specific effect in a particular segment of the media presentation data stream. In this way, specific effects may be synchronized to the media presentation.
The opponent submitted, through Dr Pieloor’s evidence at [78] and [86], that D2 disclosed the features of claims 1 and 3 of the application. Dr Pieloor referred to the discussion at [0026] of D2 of a CRC algorithm or a hash algorithm to generate an identifying value from a segment or multiple segments of data in a data stream, and that several or a sequence of values may be identified [0065], to suggest this was the same as the claimed method of determining a sequence of a plurality of hash values from in-band data corresponding to a user-selected location within a content stream. On the other hand, there is nothing in those passages of D2 of responsiveness to user input of selected locations of a content stream. That is, there is no disclosure of the bookmarking of the stream and the identification of sequences of hash values only at the bookmarks as claimed in the present application.
Dr Pieloor further indicated that D2, at [0035], stated that an effect signal is associated with the point in time that the user has selected. On the other hand, [0035] clearly relates this aspect to the creation of the above-mentioned second table. Moreover the opening statement of [0035] indicates it is an effect arranger, programmer, content provider or other user that uses the system of D2 to create the table. The expression “other user” should be read in context. It would seem appropriate to limit this expression to users like the preceding effect arranger, programmer, or content provider. The previously mentioned content enhancer might fairly be said to additionally fall within the scope of “other user”. Also, the passage is about creating effects for particular time points through the creation of the second table. The end user would not be the one creating the second table or the effects. I would conclude the end user is not with the scope of the expression “other user” in the present case.
The primary purpose of the invention in D2 is to synchronize the effects with the media presentation. There is little in D2 about user selection of specific locations in a content stream and, responsive thereto, determining a sequence of hash values and generating metadata associated with those locations. Rather, D2 appears to require the execution of the latter two steps over much of the data stream to achieve the synchronization of the specific effects with the presentation as determined by the content provider, programmer or arranger.
I conclude the claimed invention is novel over D2.
Inventive Step
In respect to inventive step, the Wellcome Foundation Ltd v VR Laboratories (Aust.) Pty Ltd decision, (1981) 148 CLR 262 at 280, states the question is whether the invention would have been obvious to a hypothetical skilled addressee armed with the common general knowledge at the priority date. Also from that decision at page 286, an appropriate test is whether a person skilled in the relevant field, and faced with the same problem, would have taken as a matter of routine whatever steps might have led from the prior art to the invention.
The High Court in Aktiebolaget Hassle v Alphapharm Pty Limited, (2002) 56 IPR 129 at [50] – [53], appeared to approve of the Wellcome (supra) test. In discussing what was meant by a matter of routine the High Court noted and accepted an affinity with the approach in Olin Mathieson Chemical Corporation v Biorex Laboratories Ltd, (1970) 87 RPC 157, of whether the person skilled in the art would directly be led as a matter of course to try what was claimed in the expectation that it might well produce a useful alternative. In Lockwood Security Products Pty Ltd v Doric Products Pty Ltd[No. 2], [2007] HCA 21, (2007) 235 ALR 202, general principles regarded to be of continuing relevance, at [50] – [52], were that “obvious” means “very plain”, a scintilla of invention remains sufficient to support the validity of a patent, there must be some difficulty overcome, some barrier to be crossed, and an invention must be beyond the skill of the calling.
Person Skilled in the Art
The opponent submitted the person skilled in the art relating to the alleged invention included an engineer familiar with multimedia content and its manipulation. I have no contrary view and accept the opponent’s statement. Dr Pieloor clearly is and had been such a person at the priority date of the application.
Common General Knowledge
At the hearing the opponent submitted, with reference to Dr Pieloor’s evidence, that all of the following items formed part of the common general knowledge in the art before the priority date.
Television data was broadcast for use by STBs using elementary streams of video and audio, and using program streams and transport streams for system coding as detailed in the AS/NZS 13818.1:2002 Standard. Subtitles or closed caption data were transmitted as in-band data. Metadata was used to describe an event for use by the STB. Metadata was used as the only practical means for instructing a device to skip a segment of a content stream.
The above statements all appear to be a reasonable account of the common general knowledge before the priority date. In the absence of evidence from the applicant, I accept the opponent’s statements.
The opponent further submitted that the FOXTEL subscription TV service enabled users to bookmark or tag specific locations in content streams. These bookmarks were stored locally on the STB for retrieval by the STB in response to user action.
In the absence of evidence from the applicant, I would consider the described bookmarking features to have been common general knowledge before the priority date. One may query whether it was common general knowledge for a specific subscription TV service to have enabled such features. On the other hand, FOXTEL has been a major and active company in the Australian subscription or pay TV market for at least 10 years before the priority date of the application. I accept the opponent’s statements in respect to the FOXTEL subscription TV service.
The opponent further submitted that hash values were used to summarize or represent a variable length data input, and that server-based distribution of files was a basic method of file sharing. I have no reason to doubt these statements and therefore I accept them.
Whether There is an Inventive Step
From the novelty discussion above, the key difference of the claimed invention over D1 and D2, individually, is the determining of hash value sequences from in-band data corresponding to user-selected locations within a content stream. One thus needs to establish whether the hash sequencing of bookmarks in the content stream would have provided the claimed invention with an inventive step before the priority date.
The object of the invention appears to be to enable a user to enhance a content stream. The title mentions creating and viewing customized multimedia segments. At [132] for example, DVR users can distribute their own sets of points of interest to other users. Users can further attach metadata to each point of interest to cause the DVR to display certain text or tell the DVR to skip parts of the content stream, for example.
I have indicated above that I consider the enabling of users to bookmark or tag specific locations in content streams to have been common general knowledge before the priority date.
Dr Pieloor stated, at [68], that hashing was well described, by the priority date, in standard computer programming texts. He cited Robert Sedgewick’s “Algorithms in C++”, [1992], Addison-Wesley Publishing Company, as a commonly referenced text. That text’s chapter 16, exhibited in evidence, is a chapter on hashing. On page 231, it is stated that hashing provides a way to use a reasonable amount of both memory and time to strike a balance between these two extremes. Efficient use of available memory and fast access to the memory are prime concerns of any hashing method. It is further stated there is a great deal of empirical and analytic evidence to support the utility of hashing for a broad variety of applications.
Dr Pieloor further stated, at [141], the computation and use of hash values for pattern matching was common and a widely known programming technique before the priority date. He cited D1 as an example. Dr Pieloor stated that the authoring computer of D1 might compute hashes of the key phrases if desired to reduce the size of the data file and improve matching speed at the client. Dr Pieloor indicated that he knew and regarded it as well known at the priority date that hashes could be used for the reasons stated in D1.
I am satisfied that, well before the priority date, it was well understood that hashing broadly involved the transformation of variable-length inputs, such as strings of characters, into fixed-length outputs, usually shorter. As such, the use of hashes for enabling reductions in file storage size requirements and enhancing processing speeds would have been well recognised. I would regard that as a mechanism that the person skilled in this art would have considered as a matter of course in applying to multimedia streams.
It then needs to be established whether it would have been obvious to take a hash sequence determination process for a content stream and apply it to only portions of the content stream at the user’s device, for example to one or more bookmarks. The desire and ability for users to bookmark multimedia streams had been well known before the priority date so as to enable some form of customization of the content or associated events at the bookmarked points by users. The specification itself opens with the statement that the present invention relates to locating points within the multimedia stream at a multimedia device, that being a user device. The attractiveness at the user’s device, in terms of use of system resources, of applying processing operations, such as hash sequence determination, only to user-selected points of a content stream would also appear to have been self-evident to the person skilled in the art before the priority date. Moreover the hash sequence determination process itself was not new, for example from D2. It is the application of the determination process only to user-selected portions of a content stream that is different. I conclude that such selective application of the known process is not inventive.
I regard claim 1 as lacking an inventive step over D2. The same applies to independent claim 3.
Dr Pieloor has analysed, at [85] – [136], the additional features of the dependent claims against D2 and what was commonly known before the priority date. In the main, I concur with Dr Pieloor’s statements. I would conclude the additional features of claims 2, 4, 6, 9-13, 15, 23, 25 and 28 are disclosed in D2 such that those claims do not have an inventive step. I would also conclude that the additional features of claims 5, 7, 8, 14 and 16-18 would have been matters of course for the person skilled in the art and thus these claims also do not have an inventive step.
Manner of Manufacture
The opponent submitted that all the features of the claims were known and therefore were below the threshold required for inventiveness and therefore the application did not claim a manner of manufacture.
It appeared the opponent’s submission related principally to lack of novelty or inventive step which has already been discussed above. On the other hand the opponent referred to the NV Philips Gloeilampenfabrieken v Mirabella International Pty Ltd case, [1995] HCA 15, 183 CLR 655, a decision on manner of manufacture. One may thus infer there was more to the opponent’s submission. The NV Philips (supra) decision considered whether there was a general threshold requirement of newness or inventiveness in S18(1)(a) of the Patents Act. The decision states that the threshold requirement of an alleged invention means that, if it is apparent on the face of the specification that the quality of inventiveness necessary for there to be a proper subject of letters patent is absent, one need go no further. In the present case there is no reference in the specification that would indicate that features claimed relating to the determination of hash value sequences from in-band data were known or not inventive.
I conclude the application relates to a manner of manufacture.
Usefulness
The opponent submitted that claims 1 and 3 lacked utility on a couple of grounds. Firstly, these claims relate to a method for creating and viewing customized multimedia streams. The opponent submitted however that neither of claims 1 and 3 included a feature that causes a customized multimedia stream to be created or viewed. The claims merely define displaying a content stream and then, after receiving input from a user, determining hash values and generating metadata and storing that information.
The Ranbaxy Australia Pty Ltd v Warner-Lambert Company LLC decision, [2008] FCAFC 82 at [141], provides guidance:
“If the claimed invention does what it is intended to do and the end obtained is itself useful, the invention is useful within the meaning of s 18(1)(c). ….the invention as claimed must attain the result promised by the patentee.”
In the present case, the opponent’s criticism appears to be directed to the absence from claims 1 and 3 of a concluding step that a customized multimedia stream is created and viewed out of the generated metadata and stored event identification data. On the other hand, the method steps leading to that outcome are defined. It would appear to be odd to regard the claimed invention as not useful only on account that the final creation and display of a customized multimedia stream is not mentioned in the claims. I think that would not be within the ambit of what is meant in the Ranbaxy (supra) case by a failure to achieve the promised result.
The opponent’s second ground for lack of utility was that claims 1 and 3 covered the sequencing of hash values from single words within closed caption text which would inevitably result in hash value sequences repeating throughout a single multimedia stream. Such repetition would not enable a user-selected location to be accurately detected.
I think a more accurate account of this criticism is that at least one user-selected location would be accurately detected but the hash sequencing from single words would not distinguish other non-selected locations from a user-selected one. This appears to be an unintended scenario of the claims. It could lead to false hits, for example of event identification data being wrongly associated with non-selected locations. In that sense the claimed invention does not do what it is intended to do. There would appear to be an easy remedy for this in claims 1 and 3 in terms of using sufficient lengths of the in-band data to create unique hash value sequences. The insertion of the word “unique”, as in “determining a unique sequence of a plurality of hash values…”, would appear to be a solution.
In respect to claim 24, the opponent submitted, with reference to Dr Pieloor’s evidence, that the consideration of when hash values are computed is not efficient or useful to the methods of claims 1 or 3. The criticism was that the specification does not explain why values representing the time of computation of hashes should be stored as part of the event identification data.
It is clear the intent here is to record or store the time differences between the computations of hash values. I am not prepared to state that is not useful. The absence of an explanation in the specification for this feature does not offend against Subsection 18(1)(c).
I conclude claims 1 and 3 do not define an invention that is useful because of the lack of uniqueness of hash value sequences that is covered by the claims as they presently stand.
Clarity and Fair Basis of Claims
The opponent, through Dr Pieloor’s evidence, mentioned a myriad of items, some of which have already been discussed above, that supposedly offended against the clarity and fair basis of the claims. I will only deal with any additional and significant items below.
Dr Pieloor stated the claimed method refers to determining a sequence of a plurality of hash values from in-band data. On the other hand, Dr Pieloor mentioned there is no description of the determination of hash values from any type of in-band data other than closed-caption data. ETV signalling data is mentioned in the specification as another form of in-band data. However, the specification states at [0127] and [0128] that, similar to recognising hash value sequences in closed-caption data, patterns of ETV signalling data are recognised to synchronize ETV events with particular portions of a video program.
The above statement from the specification may suggest that hash sequences are exclusive to closed-caption data. That is, ETV signalling does not involve the calculation or determination of hash value sequences. On the other hand, paragraph [0054] states the invention is not limited to closed-caption and ETV signalling data, and may be applied to other types of in-band data. Closed-caption data may be the only example provided that involves hashing. However the specification at [0054] is open to the hashing of other types of in-band data. The use of hashing as a matter of course has been mentioned above for data transformation to reduce file storage needs and enhance processing speeds. With ETV signalling for example, there would appear to be no barrier to hashing ETV signalling data similar to that described for closed-caption data. I conclude the broad definition of “in-band data” in claims 1 and 3 is consistent with the description and thus these claims are fairly based on this point.
In the light of the above, the opponent’s submission that claim 5 is not fairly based, because of the definition of the in-band data being ETV signalling data, is moot.
Dr Pieloor stated that claims 2 and 13 are unclear in defining the event identification data, received at the second device, associating a particular pattern with a location in the second content stream. On the other hand, the event identification data, as per claims 1 and 3, comprises a sequence of hash values and metadata. There is no mention in claim 1 or claim 3 of the storage of any pattern other than a sequence of hash values as event identification data. Moreover, in view of the definition of the second content stream containing at least a portion of the in-band data contained in the first stream, and the absence of defining a unique sequence of hash values, the ability of the second device to uniquely identify locations is an issue in the present claims 2 and 13. I conclude claims 2 and 13 are not clear.
Dr Pieloor referred to claim 12 in respect to its definition of metadata being associated with a pattern that can describe actions relating to one or more locations. He indicated there is no description in the specification of metadata associated with a pattern describing actions relating to more than one location. The thrust of the specification is that each pattern, or perhaps more clearly each hash value sequence, is associated with a single location and not multiple locations. I conclude that claim 12 is not fairly based.
Dr Pieloor stated that claims 21 and 22 do not fit together with claims 1, 2 or 13. The former claims relate to combining program segments from different content streams. However there is no description of this activity being associated with the generation and sharing of event identification data for a specific content stream, as in claim 1 for example, or how the event identification data is associated with specific segments extracted from different content streams. I conclude claims 21 and 22 are not fairly based.
Sufficiency of Description
The opponent submitted that the specification failed to describe the invention fully on several counts. These generally overlap with items already discussed above under usefulness, and clarity and fair basis of claims. In most cases, the items raised under the ground of lack of sufficiency are more appropriate under one of the above grounds. Having reviewed Dr Pieloor’s declaration, I cannot see any additional items of substance that would offend under the ground of lack of sufficiency.
I conclude the specification fully describes the alleged invention.
Conclusion
I have found that claims 1-18, 23, 25 and 28 do not have an inventive step. I have also found that claims 1 and 3 are not useful, that claims 2 and 13 are not clear and that claims 12, 21 and 22 are not fairly based.
It is clear that suitable amendments could be made to the claims to overcome the above deficiencies. I allow the applicant two (2) months from the date of this decision in which to propose suitable amendments.
Costs
The opponent sought the costs of these proceedings. Costs normally follow the event. I see no reason to depart from this practice in this case. The opposition is successful. I award costs in accordance with Schedule 8 against the applicant, TiVo, Inc.
M G Kraefft
Delegate of the Commissioner of Patents
0
8
0