JP4331706B2 - Editing apparatus and editing method - Google Patents

Description

  The present invention relates to a content playlist editing apparatus and editing method, and in particular, using a shared playlist stored on a server and selecting an appropriate playlist for a content being edited by a user, The present invention relates to an editing apparatus and an editing method for extracting information necessary for editing from a selected playlist and supporting user editing work.

  HDD (Hard Disk Drive) recorders have become widespread, and it has become possible to create and edit playlists of content as well as simply storing content recorded from television broadcasts.

  Authoring to add annotations and chapters to recorded content is performed by the user individually, and as a technology to support them, titles and comments can be used using EPG (Electronic Program Guide). It was used for recognizing and attaching video and audio and telop.

For example, in Patent Document 1, scene cut is performed using an “index” of video tape control information. An apparatus for sharing metadata describing information related to content data is disclosed in Patent Document 2.
JP-A-6-176542 (pages 3 to 4, FIG. 2) JP 2004-193871 A (pages 6-7, FIG. 1)

  However, it has not been possible to sufficiently assist the user in content editing work such as chapter editing and chapter titles.

  An object of the present invention is to provide an editing apparatus and an editing method for editing a playlist of contents.

  According to a first aspect of the present invention, there is provided a storage unit that stores content, an editing unit that edits an edit playlist of content stored by the storage unit, and a shared playlist storage device that share content stored by the storage unit A search unit that searches whether playlists are managed, a calculation unit that calculates a similarity between the edited playlist and the searched shared playlist, and a similarity when the similarity is higher than a predetermined value The editing apparatus includes a support unit that supports editing of the edited playlist by using information on the shared playlist with high frequency.

  A second invention is the editing apparatus according to the first invention, wherein the calculation unit calculates the similarity between the chapter cut of the edit playlist and the chapter cut of the shared playlist.

  A third invention is characterized in that, when the similarity of chapter cuts by the calculation unit is higher than a predetermined value, the support unit applies chapter cuts of the shared playlist to chapter cuts of the edit playlist. The editing apparatus according to the second invention.

  According to a fourth aspect of the present invention, the calculation unit calculates a similarity of a shared playlist created by the same creator as the creator of the shared playlist that has been used in the past. It is an editing apparatus of description.

  A fifth invention is the editing apparatus according to the first invention, wherein the calculation unit calculates a similarity between the chapter of the edit playlist and the chapter of the searched shared playlist.

  A sixth invention is characterized in that the content stored in the storage unit is composed of a plurality of contents, and the edit playlist and the shared playlist also correspond to the plurality of contents. The editing apparatus according to the first aspect of the invention.

  A seventh invention is the editing apparatus according to the first invention, wherein the editing playlist and the shared playlist are created in a desired talent unit.

  The eighth invention includes a step of editing a playlist of accumulated content, a step of searching whether the shared playlist storage device manages a shared playlist related to the stored content, the editing playlist, The step of calculating the similarity with the searched shared playlist, and when the similarity is higher than a predetermined value, the editing of the edited playlist is performed using the information of the shared playlist having a high similarity. And an editing step.

  According to the present invention, it is possible to provide an editing apparatus and an editing method for editing a playlist of contents.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First embodiment)
FIG. 1 is a schematic block diagram of an editing system according to the first embodiment of the present invention, which is constructed by a server client model. The server 200 includes a shared playlist storage unit 201 that stores a playlist of content that can be used by a plurality of client terminals, and a communication unit 202 that can communicate with each client terminal. On the other hand, the client terminal 100 which is an editing device includes a playlist search unit 101, a playlist similarity calculation unit 102, an editing support information extraction unit 103, a playlist editing unit 104, a playlist storage unit 105, a content A storage unit 106 and a communication unit 107 capable of communicating with the server 200 are provided. An example of the client terminal 100 is an HDD (Hard Disk Drive) recorder.

  FIG. 2 is an example of a shared playlist stored in the shared playlist storage unit 201.

  FIG. 3 is a flowchart of the entire editing system according to the first embodiment. When the user of each client terminal, that is, the playlist creator, records the broadcast content in the content storage unit 106, the playlist editing unit 104 creates the playlist of the recorded content, and the playlist storage unit It accumulates in 105 (S301). Further, the playlist is shared by transmitting the created playlist to the shared playlist storage unit 201 (S302).

  Further, the playlist editing unit 104 can receive playlist creation support by the shared playlist by searching the playlist search unit 101 for the shared playlist recorded in the shared playlist storage unit 201 (S303). Then, the playlist that received the creation support is transmitted again to the shared playlist storage unit 201 (S304), and the playlist is shared (S302).

  FIG. 4 is an example of a display screen of the playlist editing unit 104. The playlist editing unit 104 includes a content display unit 4001, a program title input / display unit 4004, a chapter title input / display unit 4002, a chapter creation button 4011, and a control for performing playback / stop / pause / fast forward / rewind of the content. Section 4003, program URI (Uniform Resource Identifier) display section 4005, timeline 4006, bar 4009 indicating the current display position, play list save button 4012, support information acquisition button 4013, and share button 4014. The control unit 4003 also includes a rewind skip button 40031, a rewind button 40032, a stop button 40033, a play button 40034, a pause button 40035, a fast forward button 40036, and a fast forward skip button 40037.

  The playlist does not have the content itself, but includes content editing information. For example, as a URI for identifying the content, the broadcast date / time / station, chapter cut position information, the entire title, and the chapter title , Comments, overall thumbnail extraction positions, and chapter thumbnail extraction positions.

  A playlist ID indicated by <playlist-id> is used to manage a plurality of playlists as a shared playlist, and is assigned when the shared playlist storage unit 201 manages the shared playlist.

  The URI for content identification indicated by <uri> is broadcast date / time / broadcast station here, but as another example, a server for identifying content is constructed, and an original ID is assigned to the content. May be distributed together with the program.

  Also, a content ID may be created from the content feature vector. The content is time-series data including image data (video data) such as still images and moving images and audio data, and the video and audio characteristic states of the content (for example, scene change, silent period, volume) The feature vector of the content can be generated on the basis of information (for example, time until each time point) indicating each time point at which the state of the level exceeds a predetermined level.

  Here, the start date is <start-year>, the start month is <start-month>, the start date is <start-date>, the start time is <start-time>, and the end time is <end-time> It is represented by

  The program title is indicated by <title>. The comment of the program is indicated by <comment>. The program title and comment may be created by a user using a playlist creation user interface as shown in FIG. 4 when creating a playlist, or may be copied from EPG (Electric Program Guide) information.

  The entire thumbnail indicated by <thumbnail> designates a certain time in the playlist. In this example, 200000 milliseconds from the beginning is specified, and the image at that time of the video is used as a thumbnail.

  The chapter cut information indicated by <chapter> is created as time information when dividing the chapters of the playlist. The chapter start time is indicated by <chapter-start-time>, and the chapter end time is indicated by <chapter-end-time>. In the example, the first chapter is from 100,000 to 100000 milliseconds from the beginning, the second chapter is from 200,000 milliseconds to 210000 milliseconds from the beginning, and the third chapter is 300,000 milliseconds from the beginning. Until 310000 milliseconds.

  The chapter title indicated by <chapter-title> can be given by the user when creating the playlist using the playlist creating user interface as shown in FIG. This allows a user to attach a chapter to each chapter created by cutting a chapter using the playlist editing unit 104.

  A chapter thumbnail indicated by <chapter-thumbnail> specifies a certain time in the playlist. In this example, the thumbnail for the first chapter is designated as 110000 milliseconds after the beginning, and the image at that time of the video is the chapter thumbnail.

  In step S303 of FIG. 3, the system presents support information for the user to create a playlist. When the user creates a playlist, a support information acquisition operation is performed so that it is not necessary to input all information for editing. This is because the support information acquisition button 4013 is pressed in FIG.

  FIG. 5 is a detailed flowchart of the playlist creation support step S303. Using the user interface as shown in FIG. 4, the user who is editing the playlist presses the support information acquisition button 4013 in order to obtain the support information of the edited playlist (S501). The playlist search unit 101 searches the shared playlist storage unit 201 for a shared playlist similar to the edited playlist (S502). This search is performed based on the URI of the editing playlist, for example. An example of the search request is shown in FIG.

  In this search request, a playlist of contents broadcast on the broadcast station from 15:00 to 17:00 on February 21, 2005 is being edited, and a shared playlist similar to that is being searched.

  At this time, the shared playlist to be searched may not be exactly the same as the search request. For example, even if the end times are not exactly the same, they are searched as being similar.

The similarity of content at this time is, for example,
Similarity to search request = (| Sq-So | + | Eq-Eo |) / Dq + Do
Can be calculated with here,
Sq: Start time of URI of search request,
So: Shared playlist start time,
Eq: End time of URI of search request,
Eo: End time of shared playlist,
Dq: URI length of search request,
Do: Length of shared playlist.

  This is obtained by normalizing the time lag with the length of each playlist to obtain the similarity. In other words, the similarity is higher if the time lag is negligible compared to the length of the playlist.

  The shared playlist set searched in this way is a set of a plurality of examples of the shared playlist shown in FIG.

  Then, the playlist similarity calculation unit 102 calculates the playlist similarity for the searched shared playlist set, and the editing support information extraction unit 103 extracts the support information (S503).

  First, the playlist similarity calculation unit 102 calculates the similarity between the shared playlist set and the shared playlist based on the similarity with the playlist being edited by the user. For example, when a certain shared playlist is created in the same manner as the playlist being created by the user, it is determined that the degree of similarity is high.

  A determination method in the case where the same creation is performed will be described by taking a calculation method based on the similarity of chapter cutting as an example.

  The user cuts chapters in order from the front of the playlist being edited. In this case, it is not necessary to cut all chapters desired by the user. For example, when the user tries to divide the playlist into three chapters, the support information acquisition operation can be performed even at the stage where the first two chapters are divided.

Playlist similarity is
(Similarity of playlist) = E / D
Can be calculated with here,
D: Number of chapter divisions in the edit playlist,
E: Number of chapter division positions in the edit playlist, the number of chapter divisions in the shared playlist within a certain time,
The certain time is a constant, for example, within 5000 milliseconds before and after.

  FIG. 7 is a conceptual diagram for comparing the playlist being edited and the shared playlist by chapter cutting. Here, it is assumed that chapters are also divided in the shared playlist (A) within a certain time at the chapter division position (C) of the edit playlist. In this case, D = 1 and E = 1, and the similarity of the playlist is 1. That is, in the playlist being edited by the user, the similarity of the playlist is defined by how much the shared playlist is divided among the positions where the user has divided the chapter, and the shared playlist and the edited playlist Are similar when chaptering is performed from the same point of view. In this way, the similarity between the playlist being edited and the shared playlist can be defined using the chapter cut position.

  A shared playlist may be created by cutting a part of the content, not the entire content. For example, assuming that the CM portion in the broadcast content is unnecessary, it can be considered that the user deletes the playlist when creating the playlist and creates the playlist with the remaining portion. Even in such a case, the similarity of content can be calculated by the above formula.

  For example, in the example of the playlist in FIG. 8, the three chapters (C1), (C2) and (C3) of the shared playlist are from 100000ms to 120000ms, 200,000 to 220000ms, and 300,000 to 320000ms of the original content. In this case, the time when the chapter is divided even in the shared playlist of the above formula is set to 100000ms, 120000ms, 200000ms, 220000ms, 300000ms, 320000ms of the start time and end time of the chapter. The similarity can be calculated using the same formula. Thereafter, when not particularly described, it is assumed that the shared playlist and the edited playlist are compared in units of chapters.

  The playlist similarity can also be calculated using the user's past usage history. Although not described in FIG. 1, the system can have a profile storage unit that stores user usage history. Although not described in FIG. 2, the shared playlist creator can also have information on the creator of the shared playlist. The profile storage unit stores information on the creator of the shared playlist used when the user edited the playlist in the past. As a result, by calculating the similarity of the shared playlist created by the same creator as the creator of the shared playlist that has been used in the past, the shared playlist creator's shared play that the user always uses is calculated. A list can be preferentially presented. The shared playlist creator information to be included in the shared playlist can give a user ID to the playlist by performing user authentication when the user accesses the system.

  From the above, the user can easily find a shared playlist similar to the playlist being edited without cutting all chapters. In this way, the shared playlist with the similarity is presented to the user in step S504. An example of the presentation screen is shown in FIG.

  In FIG. 9, the shared playlist 1 titled with the drama “XX” episode 1, the shared playlist 2 titled with the drama “XX”, and the shared titled “XX” A total of three shared playlists of the playlist 3 are presented.

  The user selects the presented shared playlist. As an example, assume that the first shared playlist is selected. Further, the type of support information is selected, and what is applied to the playlist being edited is selected from the selected shared playlist such as chapter cut, chapter title, chapter thumbnail, or the like. As a whole, it is possible to create the exact same playlist. In the screen example of FIG. 9, there are a title button, a comment button, and a chapter cut button, and the support information to be applied can be selected by pressing these buttons.

  Next, in step S505, the support information is applied to the edit playlist. The operation at this time will be described in detail by taking as an example a case where the user selects chapter cutting.

  As shown in FIG. 9, on the screen where the similarity is calculated and the shared playlist similar to the playlist being edited is presented at the top, the playlist that matches the user's intention is selected, and the editing support information is displayed. select. For example, when chapter cut is selected by pressing the chapter cut button of FIG. 9, chapter cut information is extracted from the selected shared playlist, and this chapter cut is applied to the playlist being edited. The application method will be described with reference to FIG.

  In FIG. 10, in the edit playlist, the user divides chapters (D) and (E), and in the selected shared playlist, the chapter is divided into three (A), (B), and (C). And In all chapters in the shared playlist, the edited playlist is referenced at the start time and end time of each chapter. If the chapter is divided within a certain period of time, the same part is already divided into chapters. Do nothing. If at any one of the start time and end time of a chapter in the shared playlist, chapter division is not performed within the fixed time in the edited playlist, chapter division is performed at that time in the edited playlist. .

  In the example of FIG. 10, the division of chapters (A) and (B) in the shared playlist is not performed because there is a division of (D) and (E) within a certain period of time even in the edited playlist. However, the division between chapters (B) and (C) of the shared playlist is not divided in the edit playlist within that fixed time, so this is applied and (E) is divided at the same time, Create new chapters (E) and (F). As a result, the user can select a chapter according to the intention only by selecting the presented shared playlist. Similarly, a shared playlist can be selected, and the entire title, title thumbnail, and comment can be extracted from the selected shared playlist and applied to the playlist being edited.

  In step S304, the user can transmit the created playlist to the shared playlist storage unit 201 for sharing. When the share button 4014 in FIG. 4 is pressed, the edited playlist is held in the shared playlist storage unit 201 and shared.

(Second Embodiment)
FIG. 11 is a schematic block diagram of an editing system according to the second embodiment. The difference from the first embodiment is that it further includes a chapter similarity calculation unit 111. Since other configurations are the same as those of the first embodiment, the description of the embodiment should be referred to.

  The overall processing flow of the second embodiment is the same as FIG. 3 as in the first embodiment. FIG. 12 shows a processing flow in the playlist creation support by the shared playlist in step S303. Among these, S1201 and S1202 are the same as S501 and S502, respectively. However, in S501 in the first embodiment, the user presses the support information acquisition button 4013 in FIG. 4, but every time the state of the edit playlist changes without any explicit user operation, The process may proceed to step S1202 to acquire a similar playlist, or may proceed to step S1202 when a user described later selects a chapter.

  It is assumed that a shared playlist set corresponding to the edited playlist has been acquired in steps S1201 and S1202. The subsequent operation will be described with reference to FIG.

  In step S1203, the user selects a chapter for presenting support information. For example, in FIG. 13, a chapter is selected by right-clicking with the mouse. In FIG. 13, it is assumed that the third chapter is selected.

  At this time, the editing system presents the type of support to be presented to the selected chapter. An example of this presentation is shown at 1301 in FIG. Here, chapter titles, chapter thumbnails, and chapter comments are presented. This type of support is presented based on information included in the shared playlist set similar to the edited playlist acquired in step S1202. In step S1204, the user selects information he / she wants to support from the types of support presented. Here, a case where a chapter title is selected will be described as an example.

  When the user selects a chapter title, chapter title information is extracted from the selected shared playlist, and a chapter title candidate list is presented to the user. An example of the presentation of chapter title candidates at this time is shown at 1302 in FIG.

  Next, a method for creating the chapter title candidate list will be described. In the chapter selected by the user from the playlist being edited, the chapter corresponding to the chapter is selected from the shared playlist set.

  This is possible by selecting chapters whose chapter start time and end time are similar. When there are a plurality of chapters in one shared playlist, the chapter similarity is calculated for all chapters, and the most similar chapter becomes the corresponding chapter in the shared playlist.

This chapter similarity is performed by the chapter similarity calculation unit 111 in FIG. This similarity is, for example,
Chapter similarity = (| Su-So | + | Eu-Eo |) / Du
Can be calculated with here,
Su: the start time of the chapter selected by the user in the edit playlist,
Eu: the end time of the chapter selected by the user in the edit playlist,
So: the start time of one chapter of a shared playlist,
Eo: End time of one chapter of a shared playlist,
Du: The length of the chapter selected by the user in the edit playlist.

  That is, the chapter similarity is high when the chapter times are similar. Thus, the chapter of the shared playlist corresponding to the chapter selected by the user is selected. A conceptual diagram of how to select the corresponding chapter at this time will be described with reference to FIG.

  When the user selects chapter (F) in FIG. 14, chapter (C) is selected as the corresponding chapter from this shared playlist. When the user selects (G) or (H), (D) is selected as the corresponding chapter. When the user selects (E), (A) or (B) is selected as the corresponding chapter.

  Next, the chapter title is extracted from the chapters of the selected shared playlist. In the example of the shared playlist in FIG. 2, this can be done by extracting the part with the tag <chapter-title>.

  Subsequently, the chapter title is presented to the user. The order of presentation depends on the presentation order score calculated by the following formula.

Presentation order score = α * (playlist similarity) + (1-α) (chapter similarity)
Where α is a suitable constant between 0 and 1.

Thereby, the chapter title candidate is presented to the user.

  In step S1206, when the user selects an intended title from the presented chapter title candidates, it is applied to the edit playlist. Application is possible by attaching the selected candidate to the selected chapter.

  Although the flow in FIG. 12 has been described with an example of adding a chapter title, it is also possible to add a chapter thumbnail and a chapter comment in the same manner.

  The difference between the second embodiment and the first embodiment is that the user selects a shared playlist to be applied in the first embodiment, whereas the user plays in editing in the second embodiment. This is the part for selecting chapters in the list.

(Third embodiment)
Next, a third embodiment will be described with reference to FIG. The difference between the present embodiment and the first and second embodiments is that the shared playlist stored in the shared playlist storage unit 201 is composed of a plurality of different broadcast programs. A shared playlist can be created by connecting a plurality of programs. An example of the playlist at this time is shown in FIG. The difference from FIG. 2 is that the URI is different for each chapter, and the URI is described in each chapter.

  In this case, the difference between the first and second embodiments is a part in which the playlist search unit 101 searches the shared playlist storage unit 201 for a shared playlist similar to the edited playlist.

  In the first and second embodiments, the shared playlist is searched based on the URI of the entire playlist. However, the playlist search unit 101 in this embodiment performs a comparison based on the URI for each chapter. Search for similar playlists.

  For example, in the shared playlist of FIG. 15, the playlist similarity is searched by the following formula.

Playlist similarity = M / 1 + | Co-Cu |
here,
M: Similarity of corresponding chapter
Co: Number of chapters in shared playlist
Cu: Number of chapters in the edit playlist Note that the similarity of the corresponding chapter can be obtained by the chapter similarity of the second embodiment.

  In this way, a shared playlist similar to the edited playlist is searched. Other processes are performed in the same manner as in the first and second embodiments.

  As a result, for example, it is possible to create a playlist not in units of programs but in units of favorite talent.

  An example of the playlist at this time is shown in FIG. In this example, the playlist is a collection of talented XXX's appearance programs.

  In the case where the user has created a playlist in which talented XXX's appearance programs are collected, for example, it is assumed that a playlist as shown in FIG. 17 is being edited. This is the one without the second chapter of the shared playlist of FIG. At this time, the search request is as shown in FIG. 18 as corresponding to the search request of FIG. 6 in the first and second embodiments. In this search request, according to the playlist similarity search formula, the playlist similarity of the shared playlist in FIG. 16 is highly evaluated.

  As described above, when the user creates a playlist for the appearance program about the talent XX, the playlist search unit is likely to search the shared playlist of FIG. 16 created from the same viewpoint. , Playlist editing support can be performed in accordance with the user's intention.

  The above-described embodiment is a preferable specific example of the present invention, and thus various technically preferable limitations are attached. However, the embodiments are appropriately combined and changed within a range not departing from the gist of the present invention. It goes without saying that it can be done.

1 is a schematic block diagram of an editing system according to a first embodiment. An example of a shared playlist stored in the shared playlist storage unit 201. 1 is a flowchart of an entire editing system according to a first embodiment. An example of the display screen of the play list edit part 104 which concerns on 1st Embodiment. A detailed flowchart of play list creation support step S303 according to the first embodiment. An example of a search request in shared playlist search step S502. The conceptual diagram which compares the play list in edit and a shared play list by how to cut a chapter. The conceptual diagram which compares the play list in edit and a shared play list by how to cut a chapter. An example of the presentation screen of the shared playlist in S504. The figure for demonstrating the application to the edit play list about the chapter cut information of a shared play list in S505. The schematic block diagram of the editing system which concerns on 2nd Embodiment. The detailed flowchart of play list preparation assistance step S303 which concerns on 2nd Embodiment. An example of the display screen of the play list edit part 104 which concerns on 2nd Embodiment. The conceptual diagram of how to select the chapter which concerns on 2nd Embodiment. An example of the shared play list which concerns on 3rd Embodiment. An example of the shared playlist of favorite talent units according to the third embodiment. An example of the favorite playlist unit edit playlist according to the third embodiment. An example of the search request which concerns on 3rd Embodiment.

Explanation of symbols

100 Client terminal (editing device)
101 playlist search unit 102 playlist similarity calculation unit 103 editing support information extraction unit 104 playlist editing unit 105 playlist storage unit 106 content storage unit 107, 202 communication unit 200 server 201 shared playlist storage unit

Claims (6)

A storage unit for storing content;
An editing unit for editing an edit playlist of the content stored by the storage unit;
A search unit for searching whether the shared playlist storage device manages a shared playlist related to the content stored by the storage unit;
And time information to divide the chapter of the content of the editing play list, and the calculation unit for calculating the degree of similarity between the time information to divide the chapter of the content of the shared playlist the search,
An editing apparatus comprising: a support unit that supports editing of the edited playlist by using information of a shared playlist having a high similarity when the similarity is higher than a predetermined value. If the similarity of the time information to separate chapters of the content by the calculation unit is higher than a predetermined value, the support unit divides the chapter of the content of the shared playlist said edit playlist time information to separate chapters of the content of The editing apparatus according to claim 1 , wherein the editing apparatus is applied to time information .   The editing apparatus according to claim 1, wherein the calculating unit calculates a similarity degree of a shared playlist created by the same creator as the creator of the shared playlist that has been used in the past. The content stored in the storage unit is composed of a plurality of content,
The editing apparatus according to claim 1, wherein the editing playlist and the shared playlist also correspond to the plurality of contents.   The editing apparatus according to claim 1, wherein the editing playlist and the shared playlist are created in a desired talent unit. Editing a playlist of accumulated content;
Searching whether the shared playlist storage device manages a shared playlist related to the stored content;
Calculating the degree of similarity between the time information for dividing the chapter of the content of the edited playlist and the time information for dividing the chapter of the content of the searched shared playlist;
An editing method comprising: a step of supporting editing of the editing playlist by using information of a shared playlist having a high similarity when the similarity is higher than a predetermined value.

Images