T 0655/13 18-06-2018
Téléchargement et informations complémentaires:
VIDEO ENCODING METHOD, DECODING METHOD, DEVICE THEREOF, PROGRAM THEREOF, AND STORAGE MEDIUM CONTAINS THE PROGRAM
Nippon Telegraph and Telephone Corporation
National University Corporation Nagoya University
Appealed decision sufficiently reasoned (no)
Remittal to the department of first instance and reimbursement of the appeal fee - fundamental deficiency in first instance proceedings (yes)
Summary of Facts and Submissions
I. The appeal is against the decision of the examining division refusing European patent application No. 06 843 675.7, filed as an international application with the Japan Patent Office (JPO) as receiving Office and published as international publication WO 2007/077942 A1.
II. During the international phase, the JPO as Inter-national Preliminary Examining Authority drew up an International Preliminary Report on Patentability (IPRP) written in Japanese. The IPRP refers inter alia to document 1 entitled "(See PDF)" ("Coding of Multi-Viewpoint Images"), this document not being a patent document but an article published in the technical journal "(see PDF)" ("The Trans-actions of the Institute of Electronics, Information and Communication Engineers"), Vol. J82-D-II, No. 11, pp. 1921-1929.
The examination file contains an English translation of the IPRP. In particular, the English translation of the reasoned statement with regard to novelty, inventive step and industrial applicability included in the IPRP reads as follows:
"The inventions set forth in claims 1, 4, 7, 9, and 11 to 14 do not involve an inventive step in the light of document 1 cited in the international search report. Document 1, in the portion of section 5. Test results entitled 'Encoding test 4,' discloses encoding wherein switching is carried out between two encoding schemes having a different number of parameters for parallax information (two-dimensional predictive encoding and epipolar predictive encoding), that is to say, the number of parameters for parallax is selected, and encoding of parallax information is carried out in accordance with the number of parameters. ..."
III. No translation of the above-cited document 1 (or of parts of it) is on file. However, the original document 1 in Japanese is available with the ISSN 0915-1923.
IV. In its communications dated 26 October 2011 and 26 June 2012, the examining division cited the above-mentioned document 1 as "D1" and raised an inventive step objection referring to "D1, paragraph 5, page 1925-1926, figures 7,8, also see the translation of the International Preliminary Report on Patentability".
V. In response to the examining division's objection, the applicants submitted in their letter dated 10 August 2012 that document D1 did not show switching between different parallax coding modes ("However, D1 does not disclose or suggest that in a single predictive encoding process, whether the dimension of the parallax is one or two can be selected for each of divided blocks of a target encoding image, that is, for reference images used for the current divided block").
VI. Thereafter the examining division issued the decision under appeal. The reasons for the decision may be summarised as follows:
(a) The subject-matter of claim 1 of the main and sole request differed from the multiview encoding method disclosed in document D1 in that
- the number of parameters for the parallax compensation was encoded (and signalled to the decoder); and
- the parallax mode (one- or two-dimensional parallax vector) could be changed on a per-block basis (see point 3.2 of the reasons).
As far as the individual features of the claimed subject-matter disclosed in document D1 were concerned, the examining division did not make any reference to specific passages. Instead, it again merely referred to "D1, paragraph 5, page 1925-1926, figures 7, 8" and to the translation of the IPRP. In this context, it further noted that the characterising part of claim 1 only defined selecting either epipolar or 2D parallax prediction and using the selected prediction mode for generating a predicted image. This was also disclosed in the above-mentioned section of document D1 (see point 3.1 of the reasons). The examining division did not indicate the source of the translation used for interpreting the pertinent passages of document D1.
(b) The claimed solution did not involve an inventive step since it would be obvious to the person skilled in the art to encode and signal the information about the number of parameters used for parallax compensation, as otherwise correct decoding would not be possible. Furthermore, choosing different encoding modes on a per-block basis was well known to those skilled in the art and was disclosed e.g. in document D2 (EP 1 414 245 A1) (see point 3.4 of the reasons).
(c) As regards the applicants' counter-arguments submitted in their letter dated 10 August 2012 (see point V above), the examining division referred to its statement in point 3.1 of the reasons and reiterated that document D1 did not only disclose the possibility of using either a one-dimensional or a two-dimensional parallax mode but also disclosed switching between those two modes within one single predictive encoding process (see point 3.5 of the reasons). Again, the examining division cited no specific passage as the source for its statement.
VII. The applicants appealed against the examining division's decision and with their statement of grounds of appeal submitted claims of new main and first auxiliary requests. They requested that the decision under appeal be set aside and that a European patent be granted on the basis of the claims of the new main request or of the first auxiliary request. They further requested that the appeal fee be reimbursed. They did not file a request for oral proceedings.
VIII. Claim 1 of the main request, which is identical to claim 1 of the main request underlying the decision under appeal, reads as follows:
"A video encoding method for encoding video images as a multi-viewpoint video image by using parallax compensation which performs prediction by using spatial parallax between the video images, the method comprising:
a parallax-parameter number setting step of selecting and setting a parameter number which indicates the number of parameters corresponding to the dimension of a parallax vector as parallax data used for the parallax compensation of each of blocks obtained by dividing a video image to be encoded using reference images;
a parallax-parameter number data encoding step of encoding data of the parameter number set in the parallax-parameter number setting step; and
a parallax data encoding step of encoding the parallax data corresponding to the number of parameters, wherein:
values of the parameter number settable in the parallax-parameter number setting step include:
a first parameter number which indicates a mode in which each parallax for a reference image assigned to each camera by which the reference image is obtained is generated using a one-dimensional parallax vector which indicates the position on the Epipolar line for a predetermined camera by using a single parameter; and
a second parameter number which indicates a mode in which each parallax for a reference image assigned to each camera by which the reference image is obtained is set using at least a two-dimensional parallax vector which indicates the position for a predetermined camera by using two parameters;
when the first parameter number is selected, a predicted image is generated based on the parallax for the reference image assigned to each camera, the parallax being generated using the one-dimensional parallax vector, and on pixel values of the relevant reference images, and in the parallax data encoding step, only the one-dimensional parallax vector is encoded; and
when the second parameter number is selected, a predicted image is generated based on the parallax for the reference image assigned to each camera, the parallax being generated using the two-dimensional parallax vector, and on pixel values of the relevant reference images, and in the parallax data encoding step, the two dimensional parallax vector is encoded."
IX. The appellants' arguments where relevant to the present decision may be summarised as follows:
None of the prior-art documents disclosed selecting the parallax prediction mode between "Epipolar and two dimension" as claimed. D2 disclosed selecting intraframe coding or motion-compensated interframe coding on a block-by-block basis. However, a person skilled in the art would not consider the selection of the encoding mode and the selection of the parallax prediction mode to be related to each other (see statement of grounds of appeal, page 2, first full paragraph: "the person skilled in the art does not account ... akin to each other").
Reasons for the Decision
1. The appeal complies with the requirements of Articles 106 to 108 and Rule 99 EPC and is therefore admissible.
2. Substantial procedural violation - insufficient reasoning, Rule 111(2) EPC
2.1 In the case at hand, the question of whether the subject-matter of claim 1 of the main request involves an inventive step hinges to a great extent on whether document D1 discloses the feature of switching between different parallax coding modes within one single predictive encoding process ("when the first parameter number is selected, ...; and when the second parameter number is selected, ..."). In their submissions prior to the decision under appeal, the appellants expressly rebutted the examining division's assessment, which also referred to the IPRP, that D1 disclosed this feature.
As its reasoning in the decision under appeal, the examining division merely repeated the statement made in its communications that document D1 disclosed the above-mentioned crucial feature. It cited no specific passage of D1 for this statement but referred only in a general way to a larger part of this document ("D1, paragraph 5, page 1925-1926, figures 7, 8") and to the translation of the IPRP, which is no more specific either (see point II above). Matters are further complicated by the fact that document D1 is written in Japanese and no (partial) translation is on file. The examining division also did not indicate the source of the translation used for interpreting the relevant passages of document D1 or reproduce a translation of those passages in the decision.
In the light of the above it needs to be established whether the examining division's reasoning complied with the legal requirement for reasoned decisions under Rule 111(2) EPC.
2.2 The obligation for the EPO under Rule 111(2) EPC to give sufficient reasons for its decisions is an embodiment of the fundamental principle of the right to be heard under Article 113(1) EPC. The latter requires that those involved be given an opportunity not only to present comments (on the facts and considerations pertinent to the decision) but also to have those comments considered, that is, reviewed with respect to their relevance for the decision on the matter (see Case Law of the Boards of Appeal, 8th edition 2016, III.B.2.4.1).
The reasoning given in a decision open to appeal has to enable the appellant and the board of appeal to examine whether the decision was justified or not. A decision therefore should discuss the facts, evidence and arguments which are essential to the decision in detail. It has to contain the logical chain of reasoning which led to the relevant conclusion (see Case Law of the Boards of Appeal, 8th edition 2016, III.K.4.2.1 with further references).
2.3 The extent to which the obligation to give reasons applies may vary according to the nature of the decision and must be determined in the light of the specific circumstances of the case. The criteria to be considered include the question of who bears the onus of presentation and the burden of proof for the facts relevant to the case, the parties' procedural course of action (are the statements made by the EPO department contested?) and, where appropriate, the characteristics of the publications relied on, such as their language and complexity.
2.4 Applying the above principles and considering the specific circumstances of the present case, the board finds as follows:
2.4.1 First, it must be established who bore the onus of presentation and the burden of proof with regard to the disputed feature's disclosure in the prior art.
(a) In examination proceedings, as far as issues relating to patentability requirements are concerned, the burden of proof - and consequently the onus of presentation of the relevant facts - lies initially with the examining division, which must provide evidence and facts to support its objection (see T 578/06, Reasons 21). As a rule, the applicant can dispute in a general way a general statement made by the examining division, which then in turn is obliged to raise a more detailed objection.
(b) Hence, it was incumbent upon the examining division to present facts showing why the subject-matter of claim 1 of the main and only request then on file did not involve an inventive step.
2.4.2 Next, the conditions under which the examining division fulfils the above obligation are to be examined.
(a) In order to give an applicant a fair chance to challenge the findings of the examining division, the latter should, as a rule, at least once identify where in the closest prior-art document each of the features of the claim in suit is disclosed (see e.g. the obiter dictum in T 70/02, Reasons 6).
Accordingly, the Guidelines for Examination in the EPO ("Guidelines") stipulate that the parts of a publication which are important for the decision must be cited in such a way that the conclusion drawn from this publication can be checked without difficulty, which is why reference should be made to each particular passage in the publication (see the current edition of the Guidelines, November 2017, Chapter E-X 2.6, and also the edition in force at the time of the appealed decision, June 2012, Chapter E-IX 5).
(b) The above rule may be relaxed if, for example, an applicant in its submissions prior to the decision had not contested the examining division's previous general finding that certain claimed features were disclosed in a prior-art document. It is then justifiable to assume that the applicant had accepted this finding (qui tacet consentire videtur ubi loqui potuit ac debuit). In such a situation, at least as far as these features are concerned, a general reference to a longer section of a prior-art document with no identification of specific passages may exceptionally be considered as sufficient identification of where in the prior-art document these features are disclosed.
(c) Furthermore, even if a reference to a specific location within a prior-art document were required to substantiate the disclosure of an individual feature, a general reference to a longer section of this document may exceptionally suffice if the relevant location can be readily identified. This may be the case, for example, if the publication is in one of the EPO's official languages and if the feature in question is reproduced in the same words as in the claim under examination and can therefore be immediately identified in the longer section without any interpretational effort.
(d) As regards pertinent prior-art publications written in a non-official EPO language, the legal requirement for reasoned decisions under Rule 111(2) EPC entails providing a translation of the relevant publication (or parts of it).
(i) It is true that there are no specific EPC provisions requiring a translation of a prior-art document into an official language of the EPO to be provided to an applicant.
(ii) However, to make its reasoning on the basis of such a document in a non-official language comprehensible to those conducting a later judicial review (Rule 112(2) EPC: "Decisions ... which are open to appeal"; see T 1123/04, Reasons 3.3), the examining division must provide the translation used in the examination proceedings of at least the relevant sections of the document (or even of the whole document, if this is necessary for its overall understanding) into an official language of the EPO. It cannot be assumed that the board members will be proficient in non-official EPO languages, and even if they are, their own translation might be different from that used in the examination proceedings. In the absence of the translation used by the examining division, the board is unable to examine the reasons for the decision, and in certain cases even whether the decision was justified or not (see point 2.2 above and e.g. T 1149/04, Reasons 5, as to the requirement for identifying the source of the translation of a Japanese prior-art document used by the examining division in the decision under appeal or the previous communications).
(iii) Furthermore, the translation requirement also applies if the applicant is proficient in the non-official language of the document referred to. The timely provision of a translation is the only way to ascertain whether there is a potential dispute over the correct translation of the section and to put the applicant in a position to have the possibly incorrect translation corrected (see also T 1343/12, Reasons 4.7.8 and 4.8 and Order 2, as to the requirement to examine inventive step on the basis of a certified translation of a Japanese prior-art document and not the machine translation thereof, which appears to be unclear and/or ambiguous in several relevant aspects).
(iv) Incidentally, the translation requirement is in line with the explanations in the Guidelines as to the handling of prior-art documents in a non-EPO language and translations of them at the examination stage. According to the Guidelines, obtaining a full translation of a crucial foreign-language document (or merely a translation of the relevant part of it if that can be easily identified) may be necessary if the examiner's assessment of its technical content is substantively challenged by the applicant. In such a case, a copy of the translation should be sent to the applicant (see Guidelines, November 2017, Chapter G-IV 4, and also the edition in force at the time of the appealed decision, June 2012, Chapter G-IV 4).
2.4.3 Finally, it must be ascertained whether in the case at hand the examining division's reasoning complied with the above requirements.
(a) The examining division did not refer, either in any of its communications or in the contested decision, to particular passages in document D1 (e.g. individual lines) in order to identify the exact disclosure of each individual feature of claim 1. Rather, in the decision under appeal it perfunctorily cited a whole section covering two pages ("D1, paragraph 5, page 1925-1926, figures 7, 8") as substantiation for the disclosed features. Its further reference to the translation of the IPRP does not aid in identifying the relevant (specific) passages, since the IPRP referred in the same general way to the portion entitled "Encoding test 4", which extends over two columns on page 1926 within that section of the document.
(b) Since the appellants - apart from the disputed feature specified in point 2.1 above - did not question the examining division's general statement regarding the disclosure of the majority of the features of claim 1 in document D1, the board finds such reasoning to be sufficient in that particular case.
(c) However, this does not apply to the disputed feature of switching between different parallax coding modes. The appellants explicitly contested the examining division's general assertion and, as a main line of argument, concluded from the very lack of disclosure of this crucial feature in document D1 that the claimed subject-matter involved an inventive step. It follows that the examining division should have identified a more specific passage in the publication in order to substantiate its assertion that the feature was disclosed in D1.
For this purpose, both the general reference to "paragraph 5" of document D1 and the reference to the translation of the IPRP were insufficient. Neither reference in itself is appropriate to determine exactly which specific passage of document D1 allegedly discloses the disputed feature.
(d) What is more, the examining division did not provide the translation of "paragraph 5" used to interpret the disclosure of this paragraph, which is written in Japanese. Without this translation, the board is not in a position to examine whether the decision was justified or not.
(e) The appellants - at least in this particular situation during the examination stage - also met the precondition of substantively challenging the EPO's assessment (see points 2.4.1(a) and 2.4.2(d)(iv) above). In view of the examining division's mere general reference to the relevant larger section of document D1 and of the typical difficulty of substantiating the absence of a feature as a negative fact, it was sufficient to simply contest the disclosure of the feature in question. Only if the examining division had already precisely identified the specific passage of document D1 which allegedly disclosed the disputed feature would the appellants have been compelled to further substantiate their standpoint and explain why the examining division's line of argument was incorrect.
(f) Consequently, the second exemption from the obligation to substantiate the disclosure of contested individual features (see point 2.4.2(c) above) does not apply in the case at hand since, in the absence of a translation, the relevant passage of the disputed feature's disclosure cannot be readily identified by reading the referenced larger section.
2.4.4 To summarise, by not precisely identifying the passage of document D1 which disclosed the feature in dispute and at the same time not providing a translation of at least the referenced longer section of said Japanese document, the examining division violated the appellants' right to a reasoned decision under Rule 111(2) EPC. This amounts to a substantial procedural violation.
3. Remittal to the department of first instance
3.1 Since there has been a substantial procedural violation, and in order to allow the appellants to argue their case before two instances, the board ex officio exercises its discretion under Article 11 RPBA and remits the case pursuant to Article 111(1), second sentence, EPC to the department of first instance for further prosecution, without analysing the appealed decision in its substantive aspects or taking a decision on the appellants' claim requests on file. The board sees no special reasons pursuant to Article 11 RPBA which would justify refraining from remitting the case to the department of first instance.
3.2 It will be necessary for the examining division to provide the translation of at least the referenced larger section of document D1 on which it bases its reasoning. If that section does not readily disclose the feature in question, the examining division will be required to identify the relevant passage as precisely as possible (e.g. by specifying the relevant lines or by reproducing verbatim the specific translated passage) and to set forth its interpretation thereof. This is the only way to enable the appellants (and, if necessary, the board at a later stage) to understand and verify whether the examining division has taken their arguments into account, thereby respecting their right to be heard pursuant to Article 113(1) EPC (see point 2.2 above).
4. Reimbursement of the appeal fee
In view of the above, the appeal is successful to the extent that the decision under appeal is set aside. Moreover, as a consequence of the substantial procedural violation, it was only by filing the appeal that the appellants were able to have their right to a reasoned decision and thus also their right to be heard restored. In view of this the board considers reimbursement of the appeal fee to be equitable (Rule 103(1)(a) EPC) and allows the appellants' request for reimbursement.
Order
For these reasons it is decided that:
1. The decision under appeal is set aside.
2. The case is remitted to the department of first instance for further prosecution.
3. The appeal fee is to be reimbursed.