JP2008099012A - Content reproduction system and content storage system - Google Patents

Abstract

A content reproduction system capable of easily reproducing only content desired by a user is obtained.
A metadata extraction unit 101 extracts metadata of content, and a metadata storage unit 103 stores the metadata. The metadata user interface unit 104 receives a user instruction 203 for the metadata stored in the metadata storage unit 103 and outputs the selected metadata as special playback information 204. The special playback execution unit 105 performs special playback of the content stored in the content storage unit 102 based on the special playback information 204.
[Selection] Figure 1

Description

  The present invention relates to a content reproduction system that reproduces content that matches a user's preference and a content storage system that accumulates content that matches a user's preference.

  In recent content storage / playback systems, a method of storing content by specifying using EPG (Electronic Program Guide) information or the like instead of specifying the stored content by a user has become common. Furthermore, for example, Patent Document 1 discloses a system that learns the user's preferences using these keywords and automatically records content. Further, for example, Patent Document 2 discloses a method of applying content metadata or the like, creating a content summary by applying the metadata according to a certain rule, and viewing only an important part.

Japanese Patent No. 3775937 Japanese Patent No. 3176893

  However, in the above conventional system, the collected content is enormous, and it is difficult to view all of the content within the limited viewing time of the user. Furthermore, there is a problem that the range of interest of the user widens, and content in a field that has a new interest cannot be accumulated until learning is completed. Further, the conventional content summarization technique also has a problem that a user needs to make a prior setting.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a content reproduction system and a content storage system that can easily reproduce and store only content desired by a user.

  A content playback system according to the present invention includes a metadata extraction unit that extracts metadata of content, and accepts selection input of metadata from the extracted metadata, and displays special playback information based on the selected metadata. A metadata user interface unit to output and a special reproduction execution unit to reproduce content corresponding to the selected metadata based on the special reproduction information.

  Since the content reproduction system of the present invention extracts the metadata of the content and reproduces the content based on the result of selecting the extracted metadata, it is possible to easily reproduce only the content desired by the user. it can.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing a content reproduction system according to Embodiment 1 of the present invention.
In the figure, the content playback system includes a metadata extraction unit 101, a content storage unit 102, a metadata storage unit 103, a metadata user interface (metadata UI) unit 104, and a special playback execution unit 105.

  The metadata extraction unit 101 is a functional unit that extracts metadata from content received from a broadcast wave or a communication network. Here, content refers to all data viewed by the user, such as video and music. The metadata is data (data serving as an index of information such as program contents, fields, performers, etc.) for explaining the contents. The content accumulation unit 102 is a functional unit that accumulates content, and the metadata accumulation unit 103 is a functional unit that accumulates the metadata 202 extracted by the metadata extraction unit 101. The metadata user interface unit 104 displays the metadata 202 stored in the metadata storage unit 103 on a display (not shown), and selection of metadata from the user by an input device such as a keyboard (not shown) as a user instruction 203. It has a function of receiving and outputting as special reproduction information 204. The special playback execution unit 105 is a functional unit that selects content stored in the content storage unit 102 based on the special playback information 204 output from the metadata user interface unit 104 and outputs the content as a special playback signal 205.

  Note that the content reproduction system and content storage system of the present invention are configured by a computer, and the metadata extraction unit 101, the metadata user interface unit 104, and the special reproduction execution unit 105 include software corresponding to each function, This is realized by hardware such as a CPU and a memory for executing the above. The content storage unit 102 and the metadata storage unit 103 are configured by a storage device such as a hard disk device or a semiconductor memory for storing data.

Next, the operation of the first embodiment will be described.
When arbitrary content 201 is received from a broadcast wave or a communication network, the content storage unit 102 stores such content 201. Further, when the content 201 is stored in the content storage unit 102, the metadata extraction unit 101 extracts the metadata 202. This is performed at the same time as accumulation, but the extraction may be performed using time such as when the system is idle.

  In the metadata extraction unit 101, the extraction unit in the time direction of the metadata 202 is a unit such as a frame unit or a shot unit, or an equally divided time for video. In the case of music content, extraction is performed in units of music, melody transitions in music, and evenly divided times. If the content to be extracted is a video, a color histogram, color arrangement, texture, intensity of motion, performer's face, or their temporal changes, scene change points, length of each scene, etc. are extracted. . For music, volume and frequency components are extracted. In addition, the title, performer, genre, broadcast date and time of content obtained from the above-described EPG and the like are also extracted as metadata 202.

  The metadata storage unit 103 stores the metadata 202 extracted by the metadata extraction unit 101. The metadata storage unit 103 also stores broadcast waves already stored in the content storage unit 102, metadata 202 of the content 201 received from the communication network, and metadata 202 such as an optical disc reproduced in the past. Yes. Further, the metadata storage unit 103 manages the metadata 202 such as face detection information and genre as described above. For example, the metadata 202 of the content itself such as “genre” and “title” and the metadata 202 of the scene in the content such as “face detection information” are hierarchically managed and stored.

  The metadata user interface unit 104 presents information on the metadata 202 to the user, causes the user to select metadata to be used for special reproduction based on the presented information on the metadata 202, and receives this as a user instruction 203. At this time, the user can select a plurality of metadata 202. Information of the content 201 that satisfies the selected metadata 202 is output to the special reproduction execution unit 105 as special reproduction information 204. The special playback execution unit 105 refers to the content storage unit 102 based on the special playback information 204 output from the metadata user interface unit 104 and performs special playback of the content 201. An example of selection of the metadata 202 and special reproduction will be described below.

  When the user actually views the content 201, the user selects the metadata 202. For example, assume that the user selects “genre” as the type of metadata 202. Accordingly, the metadata user interface unit 104 displays a list of genres such as “drama” and “news” based on the metadata 202 managed and stored in the metadata storage unit 103. As a result, when news is selected from the user, the metadata user interface unit 104 displays a list screen of news contents stored in the metadata storage unit 103. Furthermore, when “face detection information” is selected as metadata information from the user, a list of scenes in which people are shown, such as casters, reporters, and news characters, is displayed on the screen. If special playback is performed at this point, scenes in which a person is shown in the news content will be played back continuously. That is, in this case, since the metadata user interface unit 104 sends “metadata of a scene showing a person” as the special reproduction information 204 to the special reproduction execution unit 105, the special reproduction execution unit 105 A scene in which a person of the content 201 corresponding to the metadata 202 is shown is extracted from the content storage unit 102 and special reproduction is performed.

  Further, when a certain face is selected from here and special reproduction is executed, a scene in which the person of the face is reflected is continuously reproduced. For example, if the selected face is the main caster, the scene where the main caster introduces the main news will be played continuously, and if it is a caster in charge of sports, the sports news will be played continuously. Or, if the prime minister, the scenes of the prime minister are continuously played. In this case, as the special reproduction information 204 sent to the special reproduction execution unit 105, the metadata 202 indicating the main caster and the prime minister is sent. The special reproduction execution unit 105 uses the main caster and the Special playback of the scene where the Prime Minister is shown.

  At this time, the content to be reproduced is not necessarily independent. For example, as shown in FIG. 2, the time information of the content is used, and if it is for one week, the main news for that week is continuously reproduced. At this time, if special reproduction time is designated, special reproduction is performed in accordance with the restriction. For example, if the first 10 seconds of each scene is specified, the first 10 seconds of the scene in which the face is reflected is reproduced continuously. On the other hand, if the total reproduction time is designated as m minutes, special reproduction such as (m / number of scenes showing a face) is performed. Alternatively, it may be possible to apply a weight according to the length of the scene.

  By performing such special reproduction, it is possible to know the outline of the news of the week in a short time together with the above-mentioned main caster face information. Or if you are a prime minister, you can learn about the trend of the prime minister over the course of a week.

  In addition to this, for example, if “tourism” is selected as the genre and a scene with a strong green color is selected in the metadata of the color histogram, for example, metadata of a scene that introduces such forests and mountains that contain a lot of green 202 is given to the special reproduction execution unit 105 as special reproduction information 204. As a result, the special playback execution unit 105 continuously plays back scenes introducing forests and mountains from the content 201 stored in the content storage unit 102.

  Further, when performing special playback, the metadata user interface unit 104 sets the playback time, sorts the selected metadata 202, and performs special playback of the information of the content 201 corresponding to the sorted metadata 202. The information 204 may be sent to the special reproduction execution unit 105. Thus, the special reproduction execution unit 105 performs special reproduction so as to satisfy the special reproduction time in the sorted order.

  As described above, according to the content reproduction system of the first embodiment, the metadata extraction unit that extracts the metadata of the content, and the selection input of the metadata from the extracted metadata are received, and the selected metadata Since it has a metadata user interface unit that outputs special reproduction information based on data and a special reproduction execution unit that reproduces content corresponding to the selected metadata based on special reproduction information, only simple operations are possible. Thus, only the content desired by the user can be easily reproduced.

Embodiment 2. FIG.
FIG. 3 is a block diagram of the content reproduction system of the second embodiment.
The illustrated content reproduction system includes a metadata extraction unit 101, a content accumulation unit 102, a metadata accumulation unit 103, a user preference learning unit 106, a preference data accumulation unit 107, and a special reproduction execution unit 108, and a user preference learning unit. 106, the preference data storage unit 107 and the special reproduction execution unit 108 constitute a user preference learning unit. The basic configuration of the metadata extraction unit 101 and the metadata storage unit 103 is the same as that of the first embodiment, but time information 206 for taking correspondence with user preference data learned by the user preference learning unit 106 is provided. The time information input to the metadata extraction unit 101 is included in the metadata 202. Note that the content storage unit 102 has the same configuration as that of the first embodiment, and a description thereof is omitted here.

  The user preference learning unit 106 is a functional unit that learns user preferences based on user information, history information 207 such as accumulation history and viewing history, and time information 206. The preference data accumulation unit 107 accumulates user preference data (a) 208 learned by the user preference learning unit 106, accumulates user preference data (b) 209 output from the special reproduction execution unit 108, and The user preference data (a) and (b) 210 is output to the special reproduction execution unit 108. Here, the user preference data (a) 208 specifies an operation history such as accumulation / playback / fast-forward by the user and metadata (genre, title, face detection information, color information, etc.) when each operation is performed. Data including data (time information, pointer, etc.). User preference data (b) 209 is metadata itself and user preferences (information indicating whether the metadata is preferred). User preference data (a) and (b) 210 are user preferences. The stored data includes only data (a) 208 or stored data including user preference data (a) 208 and user preference data (b) 209.

  The special playback execution unit 108 updates the user preference data based on the metadata 202 stored in the content storage unit 102 and the user preference data (a) and (b) 210 obtained from the preference data storage unit 107, The updated user preference data is output to the preference data storage unit 107 as user preference data (b) 209. Also, the special reproduction execution unit 108 proposes to the user that special reproduction is performed when representative metadata reflecting the user's preference is specified (outputs a proposal 211 to the user), and As a result, when a special reproduction instruction 212 is input from the user, a content special reproduction signal 213 is output.

Next, the operation of the second embodiment will be described.
The metadata extraction unit 101 extracts content metadata 202. As an extraction unit in the time direction, in the case of video, a unit such as a frame unit or a shot unit, or an equally divided time is used. In the case of music content, extraction is performed in units of music, melody transitions in music, and evenly divided times.

  Also, if the content to be extracted is a video, the color histogram, color arrangement, texture, intensity of movement, performer's face, or their temporal change, scene change point, length of each scene, etc. Extract. For music, volume, frequency components, etc. are extracted. In addition, the title, performer, genre, broadcast date and time of content obtained from EPG and the like are extracted as metadata.

  The content storage unit 102 stores content 201 received from broadcast waves, communication networks, and the like. The metadata storage unit 103 stores the metadata 202 extracted by the metadata extraction unit 101. The metadata storage unit 103 also stores broadcast waves already stored in the content storage unit 102, content metadata received from the communication network, and optical disk metadata reproduced in the past.

  The user preference learning unit 106 learns user preferences based on history information 207 indicating user operations, content 201 accumulation history, and viewing history. The operation by the user who learns is a part skipped or fast forwarded, a part reproduced repeatedly, etc. when viewing the content 201. That is, the skipped part or the fast-forwarded part is determined to be a part that does not meet the user's preference, and the repeatedly reproduced part is determined to be a part that matches the user's preference. Further, when connected to a computer network such as the Internet, a user may learn a search key used for information search. Further, the user preference may be learned based on the genre of the content that the user has instructed to store, the genre and the number of times of the content that has been viewed. These learning results are output to the preference data storage unit 107 as user preference data (a) 208. When there are a plurality of users, the user preference learning unit 106 performs learning for each of those users.

  The preference data accumulation unit 107 accumulates user preference data (a) 208 output from the user preference learning unit 106. The special playback execution unit 108 performs special playback of the content 202 of the content storage unit 102 while referring to the metadata 202 of the metadata storage unit 103 and the user preference data (a) and (b) 210 of the preference data storage unit 107. To do. Examples of this special reproduction are as shown in the following (1) to (5).

(1) Assume that the user has accumulated news content for a certain day, and has normally watched a specific part, for example, economic news, and fast-forwarded watching other parts. Such operations and histories are input to the user preference learning unit 106 and output from the user preference learning unit 106 as user preference data (a) 208.
(2) User preference data (a) 208 output from the user preference learning unit 106 is accumulated in the preference data accumulation unit 107. For example, stored content genre → news, normal viewing scene (economic news) → data specifying metadata at that time, fast-forward viewing scene (part other than economic news) → data specifying metadata at that time, etc. is there.
(3) The operation of (2) is repeated for several days.

(4) The special reproduction execution unit 108 accumulates the similarity of the metadata 202 corresponding to the user preference data (a) 208 for several days accumulated in the preference data accumulation unit 107 in the metadata accumulation unit 103. The metadata 202 is compared. The metadata 202 of the portion that is normally viewed normally is compared, and if they are similar, the representative metadata of the portion that is normally viewed is created. As a creation method, various methods such as a method of performing statistical processing such as an average value of the metadata 202, a first day, or a last day are applicable. Similarly, for the fast-forward viewing, the representative metadata 202 is accumulated.

  For example, in the case of the same news program, the responsible person is often determined for each caster. In the present embodiment, the metadata 202 of the content 201 is accumulated in units of frames or shots, and the user preference data (a) 208 knows where to fast-forward, so that the user is fast-forwarding every day. Metadata, in this case the caster's face, voice and screen composition, subtitle text information, etc. other than the economic news, and the metadata that is played normally every day, in this case the economic news caster Metadata such as face, voice, screen composition, subtitle text information can be specified.

  When the metadata reflecting the user's preference is specified, the special playback execution unit 108 outputs the user preference data (b) 209 to the preference data storage unit 107. That is, the user preference data (b) 209 is data including metadata itself and user preferences (whether or not the metadata is preferred).

(5) When the representative metadata is specified in (4), the special playback execution unit 108 specifies the scene of the content corresponding to the representative metadata, and notifies the user of special playback. Proposal 211 is made. For example, the proposal 211 to the user is a process of displaying contents for performing special reproduction of content corresponding to representative metadata. Thereby, when there is a special reproduction instruction 212 from the user, the corresponding content 201 is taken out from the content storage unit 102 and special reproduction is started (a content special reproduction signal 213 is output).

  After the user preference data (b) 209 is accumulated in the preference data storage unit 107, user preference data (a) and (b) including user preference data (a) 208 and user preference data (b) 209 are stored. 210 is output to the special reproduction execution unit 108, and the special reproduction execution unit 108 updates the user preference data (b) 209 based on the user preference data (a) and (b) 210.

  In this way, by acquiring metadata and learning from the user's operation every day, it is possible to specify only the part that the user is watching every day, in this case only economic news. It is possible to view only a desired portion without instructing fast-forwarding.

  In addition to the economic news listed here, only sports news is played, only the part where a specific actor appears in the drama content, only the specific singer appears in the music content, etc. Special playback is possible.

  As described above, according to the content reproduction system of the second embodiment, the metadata extraction unit that extracts the metadata of content, and the user based on the user's operation among the metadata extracted by the metadata extraction unit. User preference learning means for obtaining metadata reflecting user's preference as user preference data, and a special playback execution unit that uses the user preference data and plays back content corresponding to the metadata of user preference data. The content reflecting the user's preference can be reproduced without performing a special operation.

Embodiment 3 FIG.
FIG. 4 is a block diagram of the content reproduction system according to the third embodiment.
The illustrated content reproduction system includes a metadata extraction unit 101, a content storage unit 102, a metadata storage unit 103, a user preference learning unit 106, a preference data storage unit 107, a preference data creation / update unit 109, a metadata search unit 110, A special reproduction execution unit 111 is provided, and the user preference learning unit 106, the preference data storage unit 107, and the preference data creation / update unit 109 constitute a user preference learning unit. Here, since the metadata extraction unit 101 to the preference data storage unit 107 are the same as those in the second embodiment, description thereof is omitted here.

  The preference data creation / update unit 109 creates user preference data (b) 209 based on the user preference data (a) 208 and user preference data (a) and (b) 210 stored in the preference data storage unit 107. And a functional unit for performing the update. That is, the preference data creation / update unit 109 is a functional unit equivalent to the user preference data creation / update function in the special playback execution unit 108 of the second embodiment. When the search key 214 is given by the user, the metadata search unit 110 receives the search key 214, user preference data (a) and (b) 210 stored in the preference data storage unit 107, and metadata storage. It has a function of collating with the metadata 202 stored in the unit 103 and outputting a search result 215. That is, the metadata search unit 110 searches the metadata stored in the metadata storage unit 103 for metadata corresponding to a given search key, and uses the searched metadata for user preference data (a), ( b) Sorting is performed based on the user preference indicated by 210, and the sorted metadata is output as a search result 215.

  The special playback execution unit 111 presents the search result from the metadata search unit 110 to the user (outputs the search result display signal 216), and when there is a special playback instruction 217 from the user, Based on this, special reproduction is performed (content special reproduction signal 218 is output).

Next, the operation of the third embodiment will be described.
The metadata extraction unit 101 extracts the metadata 202 of the content 201. The content storage unit 102 stores content received from broadcast waves, communication networks, and the like. The metadata storage unit 103 stores the metadata 202 extracted by the metadata extraction unit 101. The user preference learning unit 106 learns user operations, content accumulation history, and viewing history, and the preference data accumulation unit 107 accumulates user preference data (a) 208 output from the user preference learning unit 106. The above operation is the same as in the second embodiment.

  In the preference data creation / update unit 109, the user preference is reflected based on the user preference data (a) 208 stored in the preference data storage unit 107 and the metadata 202 stored in the metadata storage unit 103. User preference data (b) 209 is created and output to the preference data storage unit 107. The preference data storage unit 107 outputs user preference data (a) and (b) 210 including user preference data (a) 208 and user preference data (b) 209 to the metadata search unit 110.

  The metadata search unit 110 first acquires a search key 214 from the user. At this time, the assumed search key 214 includes, in addition to text, an image itself, a face portion of a person in the image, a song itself, a part of its melody, and the like.

  This search will be described below. For example, in the situation where the user is currently viewing the content, if the user performs some operation, the entire reproduced screen or a part thereof, or the reproduced music is used as the search key 214. Is assumed. Here, for example, a case where a person who is viewing the content is displayed and the face of the person is used as the search key 214 will be described as an example.

  In this case, the search results are only frames or shots in which the person's face is shown. At this time, if only the face metadata is accumulated, the face is sorted in the matching order or the time series order. However, since the user preferences are accumulated at the same time, when the user designates sorting, the user preferences can be sorted in the order of the user's preference order, the viewing frequency order, or the reverse order thereof. The order of viewing times is the order of the number of times the user has viewed content or a specific scene, and the order of preference is the stored user preference data (a) and (b) 210, The order is determined by comparing the metadata 202 of the scene in the content 201. As the order of preference, for example, the preference of a certain content is obtained by calculation such as (content preference = number of times of viewing × time of normal viewing).

  Thus, for example, it is assumed that there are contents A and B having a length of 30 minutes, and the contents A is viewed 10 times, but is almost fast-forwarded and only viewed about 5 minutes at a time. On the other hand, if the content B is viewed only three times but is viewed all 30 minutes per time, it is determined that the preference order of the content B is higher. Or, if the user is always watching the content of a specific actor and has accumulated new content 201 that has recently been aired, even if the user has not watched the content 201 even once, the content order is somewhat higher In some cases, a determination result may be obtained such as being ranked.

  The special playback execution unit 111 outputs a search result display signal 216 based on the search result 215 output from the metadata search unit 110. Thereby, when there is a special reproduction instruction 217 from the user, a content special reproduction signal 218 is output. In this way, by sorting in order of user preference, viewing frequency, or vice versa, the user can quickly search for favorite scenes that have been viewed many times or search for scenes that have never been viewed. It becomes possible.

  As described above, according to the content reproduction system of the third embodiment, the metadata extraction unit that extracts the metadata of content, and the user based on the user's operation among the metadata extracted by the metadata extraction unit. The user preference learning means for obtaining metadata reflecting user's preference as user preference data and the metadata corresponding to the given search key are searched from the metadata extracted by the metadata extraction unit. A metadata search unit that rearranges data based on user preferences indicated by user preference data, outputs the rearranged metadata as a search result, and a special playback execution unit that plays back content based on the search result Therefore, it is possible to reproduce the content reflecting the user preference in the order specified by the user.

  Further, according to the content reproduction system of the second and third embodiments, the user preference learning means obtains the user preference data by using the part where the reproduction of the content is fast-forwarded. Therefore, it is not necessary for the user to perform a special operation, and user preference data can be easily obtained.

  Further, according to the content reproduction system of the second and third embodiments, the user preference learning means obtains the user preference data by using the portion where the reproduction of the content is skipped. Therefore, it is not necessary for the user to perform a special operation, and user preference data can be easily obtained.

  Further, according to the content reproduction system of the second and third embodiments, the user preference learning means obtains the user preference data by using the portion where the reproduction of the content is repeatedly performed. It is not necessary for the user to perform a special operation for learning, and user preference data can be easily obtained.

  In addition, according to the content playback system of Embodiments 1 to 3, the content storage unit that stores content is provided, and the special playback execution unit executes special playback based on the content stored in the content storage unit Therefore, in the system for accumulating content, only the content desired by the user can be reproduced.

Embodiment 4 FIG.
FIG. 5 is a block diagram of the content storage system according to the fourth embodiment.
The illustrated content storage system includes a metadata extraction unit 101, a content storage unit 102a, a metadata storage unit 103a, a user data storage unit 112, and a content storage determination unit 113. The metadata extraction unit 101 has the same function as the metadata extraction unit 101 in each of the above embodiments. The content accumulation unit 102 a is configured to accumulate only the content based on the accumulation determination result 219 output from the content accumulation determination unit 113 and the user instruction 220. The metadata storage unit 103a is a storage unit that stores metadata corresponding to the content actually stored in the content storage unit 102a based on the storage result 221 output from the content storage unit 102a.

  The user data storage unit 112 is a functional unit that learns user preferences based on the user instruction 220, creates user preference data, and stores the user preference data. That is, the user data storage unit 112 corresponds to the function of the user preference learning unit including the user preference learning unit 106, the preference data storage unit 107, and the preference data creation / update unit 109 in the third embodiment. Data (a) and (b) 210 are output to the content accumulation determination unit 113. The content accumulation determination unit 113 reflects the user preference based on the reference metadata 222 given from the metadata accumulation unit 103a and the user preference data (a) and (b) 210 outputted from the user data accumulation unit 112. This is a functional unit that outputs the content accumulation determination result 219 to the content accumulation unit 102a.

Next, the operation of the fourth embodiment will be described.
The metadata extraction unit 101 extracts the metadata of the content 201 and outputs it as extracted metadata 202 to the content accumulation determination unit 113. The extracted metadata 202 is metadata corresponding to all the contents, and the reference metadata 222 indicates the metadata of the contents stored in the contents storage unit 102a.

  The content accumulation unit 102a accumulates content received from broadcast waves, communication networks, and the like. In the fourth embodiment, only the content determined to be accumulated is accumulated based on the accumulation determination result 219 by the content accumulation determination unit 113. . The metadata storage unit 103a stores only the metadata corresponding to the content corresponding to the content actually stored in the content storage unit 102a out of the extracted metadata 202 output from the metadata extraction unit 101. The data is accumulated based on the accumulation result 221 output from the unit 102a.

  Further, the user data storage unit 112 stores user preference data based on a content storage instruction or a content playback instruction by the user included in the user instruction 220, and the user preference data ( a) and (b) 210 are output. The content accumulation determination unit 113 reflects user preferences based on the user preference data (a) and (b) 210 given from the user data accumulation unit 112 and the reference metadata 222 given from the metadata accumulation unit 103a. The accumulation determination result 219 is output to the content accumulation unit 102a so as to accumulate only the content. That is, the accumulation determination is performed based on whether the currently received content 201 matches the user preference according to the user preference data (a) and (b) 210 and is similar to the reference metadata 222. Based on the accumulation determination result 219 output from the content accumulation determination unit 113, the content accumulation unit 102a automatically accumulates only the content that meets the user's preference among the content 201.

  Here, since the user's preference is hardly learned at the initial stage of content accumulation in the content accumulation unit 102a, the metadata accumulated in the metadata accumulation unit 103a does not necessarily reflect the user's preference. . Thereafter, the user preference is learned by the user data storage unit 112 based on the user instruction 220, and the learning result is accumulated. As a result, out of the reference metadata 222 given from the metadata accumulating unit 103a, only metadata that matches the user's preference is selected, and an accumulation determination result 219 is output so that content corresponding to this metadata is accumulated. Is done.

  In addition, even if the content is stored in the content storage unit 102a according to the user instruction 220, the user preference data (a) and (b) 210 corresponding to such content is displayed when the user hardly views the content thereafter. Since it is not output, by continuing to accumulate, only the content that more reflects the user's preference is accumulated in the content accumulation unit 102a. On the other hand, when content in a new field that the user has not been interested in until now is accumulated and reproduced (when such a user instruction 220 for accumulation and reproduction of such content is given), such content Is stored in the metadata storage unit 103a, and as the user preference is learned in the user data storage unit 112, content reflecting such user preference is stored in the content storage unit 102a. It will be accumulated.

  For example, conventionally, accumulation by a keyword by a user is mainly performed, and when the keyword “baseball” is designated, only a large amount of baseball content is accumulated in the content accumulation unit 102a. On the other hand, in the fourth embodiment, the user does not accumulate the game of other teams by utilizing the user preference, such as when the user is mainly viewing the game of a specific team. Content can be accumulated. Furthermore, by using metadata such as face information, it is possible to narrow down to a specific player's participation scene.

As described above, in the fourth embodiment, it is possible to store contents that match user preferences, and to reduce the storage capacity.
In the fourth embodiment, the case where the metadata is similar and the user preference matches is the accumulation target. However, all content may be stored and used to narrow down the priority deletion target when the content storage capacity is insufficient. Alternatively, when storing content, a variable bit rate method may be adopted and used to determine the amount of bits to be allocated. In other words, bits may be allocated according to the user's preference and compressed and stored.

  As described above, according to the content storage system of the fourth embodiment, a metadata extraction unit that extracts content metadata and a metadata storage unit that outputs metadata corresponding to the stored content as reference metadata. And a user data storage unit that stores user preference data indicating metadata reflecting the user's preference, and the metadata extracted by the metadata extraction unit matches the metadata of the user preference data, and A content accumulation determination unit that obtains metadata that matches or is similar to the reference metadata and outputs an accumulation determination result for content corresponding to the metadata, and a content accumulation unit that accumulates content based on the accumulation determination result; It is possible to store content that reflects user preferences and store It becomes possible to reduce the content to be.

Embodiment 5. FIG.
FIG. 6 is a block diagram of the content storage system according to the fifth embodiment.
The illustrated content storage system includes a metadata extraction unit 101, a content storage unit 102a, a metadata storage unit 103a, a content storage determination unit 113a, and a metadata user interface (metadata UI) unit 114. Here, since the metadata extraction unit 101, the content storage unit 102a, and the metadata storage unit 103a have the same configuration as in the fourth embodiment, description thereof is omitted here. The content accumulation determination unit 113 a is configured to perform content accumulation determination based on the accumulation target metadata 223 output from the metadata user interface unit 114 and to output an accumulation determination result 219. That is, it has a function of instructing the content storage unit 102a to store only the content corresponding to the metadata that matches the storage target metadata 223 among the extracted metadata 202 of the content 201 that is currently received.

  The metadata user interface unit 114 presents the reference metadata 222 output from the metadata storage unit 103a to the user, receives a user instruction 220 based on the reference metadata 222, and stores the metadata selected in the user instruction 220 as the storage target metadata. The data 223 is output to the content accumulation determination unit 113a.

Next, the operation of the fifth embodiment will be described.
The metadata extraction unit 101 extracts the metadata of the content 201 and outputs it as extracted metadata 202 to the content accumulation determination unit 113a. The content accumulation unit 102a accumulates content received from a broadcast wave, a communication network, or the like based on the accumulation determination result 219 output from the content accumulation determination unit 113a. The metadata storage unit 103a stores only the metadata corresponding to the content corresponding to the content actually stored in the content storage unit 102a out of the extracted metadata 202 output from the metadata extraction unit 101. The data is accumulated based on the accumulation result 221 output from the unit 102a.

  The metadata user interface unit 114 presents the reference metadata 222 output from the metadata storage unit 103a to the user, receives the user instruction 220 indicating a selection instruction from the user, and outputs the storage target metadata 223. To do. That is, the accumulation target metadata 223 is metadata designated as a accumulation target by the user in the reference metadata 222. In the metadata user interface unit 114, the user can select a plurality of metadata.

  Based on the accumulation target metadata 223, the content accumulation determination unit 113a accumulates only the content of metadata that matches all or part of the accumulation target metadata 223 among the extracted metadata 202 of the currently received content 201. The accumulation determination result 219 is output to the content accumulation unit 102a so as to be targeted. As a result, the content storage unit 102a automatically stores only the content that meets the user's preference among the content 201 based on the storage determination result 219 output from the content storage determination unit 113a.

  For example, suppose that a user likes a specific actor and wants to watch other appearances while playing movie content stored on an optical disc. At this time, if the metadata user interface unit 114 designates the actor's face information as the accumulation target as the metadata, the content accumulation determination unit 113a thereafter accumulates the content showing a similar face as the accumulation target. The determination result 219 is output to the content storage unit 102a.

  In addition, when content in a new field that the user has not been interested in until now is stored and played back (when the user instruction 220 for storing and playing back such content is given), this is the beginning. The metadata corresponding to the content is not stored in the metadata storage unit 103a, but as the user instruction 220 is given to the content storage unit 102a, the metadata corresponding to such content is stored in the metadata storage unit. 103a is accumulated. Then, in the metadata user interface unit 114, when the user selects the metadata of such content, the content reflecting the new user preference is accumulated.

  As described above, according to the content storage system of the fifth embodiment, a metadata extraction unit that extracts content metadata and a metadata storage unit that outputs metadata corresponding to the stored content as reference metadata. And a metadata user interface unit that accepts user selection for reference metadata and outputs storage target metadata information indicating the metadata of the selection result, and a storage target for the metadata extracted by the metadata extraction unit A content accumulation determination unit that obtains metadata that matches or is similar to the metadata of the metadata information, outputs a storage determination result that targets content corresponding to the metadata, and stores content based on the storage determination result Content that reflects the user's preferences Both can be performed accumulation, the user becomes possible to increase the content to be easily accumulated target, thereby improving the convenience.

  In addition, according to the content storage systems of the fourth and fifth embodiments, the content storage unit stores content when receiving an instruction to store content by the user. Even when content in a new field that does not indicate is stored, content reflecting such user preferences can be stored.

BRIEF DESCRIPTION OF THE DRAWINGS It is a block diagram which shows the content reproduction system by Embodiment 1 of this invention. It is explanatory drawing which shows operation | movement of the content reproduction system by Embodiment 1 of this invention. It is a block diagram which shows the content reproduction system by Embodiment 2 of this invention. It is a block diagram which shows the content reproduction system by Embodiment 3 of this invention. It is a block diagram which shows the content storage system by Embodiment 4 of this invention. It is a block diagram which shows the content storage system by Embodiment 5 of this invention.

Explanation of symbols

  101 metadata extraction unit, 102, 102a content storage unit, 103, 103a metadata storage unit, 104, 114 metadata user interface unit, 105, 108, 111 special playback execution unit, 106 user preference learning unit, 107 preference data storage 109, preference data creation / update unit, 110 metadata search unit, 112 user data storage unit, 113, 113a content accumulation determination unit, 201 content, 202 metadata, 203, 220 user instruction, 204 special playback information, 205 special Playback signal, 206 time information, 207 history information, 208 user preference data (a), 209 user preference data (b), 210 user preference data (a), (b), 211 suggestions to users, 212 specials from users Playback instruction, 213 , 218 Content special playback signal, 214 search key, 215 search result, 216 search result display signal, 217 special playback instruction from user, 219 storage determination result, 221 storage result, 222 reference metadata, 223 storage target metadata.

Claims (10)

A metadata extractor for extracting content metadata;
A metadata user interface unit that accepts metadata selection input from the extracted metadata and outputs special reproduction information based on the selected metadata;
A content reproduction system comprising a special reproduction execution unit that reproduces content corresponding to selected metadata based on the special reproduction information. A metadata extractor for extracting content metadata;
User preference learning means for obtaining metadata reflecting user's preference as user preference data based on the user's operation among the metadata extracted by the metadata extraction unit;
A content reproduction system including a special reproduction execution unit that reproduces content corresponding to metadata of the user preference data using the user preference data. A metadata extractor for extracting content metadata;
User preference learning means for obtaining metadata reflecting user's preference as user preference data based on the user's operation among the metadata extracted by the metadata extraction unit;
Searching for metadata corresponding to a given search key from the metadata extracted by the metadata extraction unit, and rearranging the searched metadata based on the user preference indicated by the user preference data, A metadata search unit for outputting the sorted metadata as a search result;
A content reproduction system comprising: a special reproduction execution unit that reproduces content based on the search result.   4. The content reproduction system according to claim 2, wherein the user preference learning means obtains user preference data using a location where the reproduction of the content is fast-forwarded.   4. The content reproduction system according to claim 2, wherein the user preference learning unit obtains user preference data using a portion where reproduction of content is skipped.   4. The content reproduction system according to claim 2, wherein the user preference learning unit obtains user preference data using a portion where the reproduction of the content is repeatedly performed.   The content storage unit for storing the content is provided, and the special playback execution unit executes the special playback based on the content stored in the content storage unit. 2. The content reproduction system according to item 1. A metadata extractor for extracting content metadata;
A metadata storage unit that outputs metadata corresponding to the stored content as reference metadata;
A user data storage unit for storing user preference data indicating metadata reflecting user preferences;
For metadata extracted by the metadata extraction unit, metadata that matches the metadata of the user preference data and matches or is similar to the reference metadata is obtained, and content corresponding to the metadata is obtained. A content accumulation determination unit that outputs an accumulation determination result to be accumulated;
A content storage system comprising: a content storage unit that stores content based on the storage determination result. A metadata extractor for extracting content metadata;
A metadata storage unit that outputs metadata corresponding to the stored content as reference metadata;
A metadata user interface unit that accepts a user's selection for the reference metadata and outputs storage target metadata information indicating metadata of a selection result;
For the metadata extracted by the metadata extraction unit, metadata that matches or is similar to the metadata of the accumulation target metadata information is obtained, and an accumulation determination result for which content corresponding to the metadata is to be accumulated is obtained. A content accumulation determination unit to output;
A content storage system comprising: a content storage unit that stores content based on the storage determination result.   10. The content storage system according to claim 8, wherein the content storage unit stores the content when receiving an instruction to store the content by the user.

Images