Hold back and real time ranking of results in streaming matching system

0001Cross-reference of related applications This application is the US Patent Application No. 1 entitled "HOLD BACK AND REAL TIME RANKING OF RESULTS IN A STREAMING MATCHING SYSTEM" filed on December 31, 2012, which is incorporated herein by reference in its entirety. Claim the interests of No. 13 / 732,108.

0002The present disclosure generally relates to matching objects in a rank-based streaming matching system, specifically holding back a match and continuously ranking the results. , Regarding giving more desirable reliability to the results.

0003Certain types of matching are based on matching a melody or audio content (eg, a song) that can be stored in a large data store with multiple different versions, or with related copies that have some differences. Such types of systems are commonly referred to as audio matching systems. In some cases, matching is particularly difficult due to differences in melodies or differences in some features of the analyzed composition. For example, a data store is a nearly perfect match due to multiple different versions with altered content, such as EP tracks and original song tracks, express songs and non-exclusive songs. Can have. When the user grants access to the audio content and the server matches the content without receiving all of the audio files, the data store holds multiple nearly exact matches. It can be difficult to match things that are exactly the same.

<p num="0004"> A brief overview of the various aspects of the disclosure is provided below to provide a basic understanding of such aspects. This overview is not a detailed overview of all possible embodiments and is not intended to identify key or significant elements or to elaborate on the scope of such embodiments. It is intended to present some of the concepts of the present disclosure in a concise form as a preparatory step to the more detailed description described below.</p><p num="0005"> The systems and methods described herein relate to matching comparison objects with multiple reference objects to identify matching results. The results are retained and updated until additional quantities of comparison objects are processed. Results are scored and with increased reliability of which results have passed the score threshold and can be output as a result prior to others, released immediately, and / or results. Compared to determine if it can be released.</p><p num="0006"> To give an example of an embodiment, the method comprises using a processor to execute a computer executable instruction stored on a computer readable medium to perform processing. The process includes receiving a probe audio sample and comparing the probe audio sample with multiple reference audio samples to identify at least one matching reference audio sample. Depending on identifying multiple matching reference audio samples, the process further includes assigning each of the ranking scores to the matching reference audio samples.</p><p num="0007"> Another example of the embodiment includes a system comprising a memory for storing computer executable components and a processor for executing computer executable components stored in micromemory. Computer-executable components include receiving components that receive the first piece of audio streaming content. The comparison component compares the first part of audio streaming content with multiple reference audio samples. The matching component identifies a matching reference audio sample that has a thing that matches well with the first part of the audio streaming content based on the comparison. The ranking component builds a first ranking score for a matching reference audio sample out of a plurality of reference audio samples.</p><p num="0008"> Also described herein are computer-readable storage media that include computer-executable instructions that cause a computing system with a processor to perform processing in response to execution. The process involves receiving a first portion of the probe audio sample via the processor. The process further includes a step of comparing the first part with the plurality of reference audio samples in order to identify the plurality of matching reference audio samples. Identify multiple matching reference audio samples that meet the well-matched mono threshold. The score of the matching reference audio sample is determined according to the set of parameters, and the matching reference audio sample that does not meet the score threshold is retained after being output.</p><p num="0009"> The following description and accompanying drawings describe in detail certain embodiments of the present disclosure. Although presenting such an embodiment, it is only a part of various methods in which the principles of the present disclosure can be used. The present disclosure is intended to include all such embodiments and their equivalents. Considered in conjunction with the drawings, other advantages and unique features of the present disclosure will become apparent in the following detailed description of the present disclosure.</p>

0010<figref num="1">It is a block diagram which shows an example of an embodiment which is not intended to limit a collation system by various aspects and embodiments described in this specification.</figref><figref num="2">It is a block diagram which shows an example of an embodiment which is not intended to limit a system by various aspects and embodiments described herein.</figref><figref num="3">FIG. 5 is a block diagram illustrating an example of an embodiment not intended to limit a system having matching components according to the various aspects and embodiments described herein.</figref><figref num="4">FIG. 5 is a block diagram illustrating an example of an embodiment not intended to limit a system having matching components according to the various aspects and embodiments described herein.</figref><figref num="5">It is a figure which shows an example of the ranking component by various aspects and embodiments described in this specification.</figref><figref num="6">FIG. 5 is an exemplary flow diagram illustrating an example of an embodiment not intended to be limiting for processing media content matching according to the various aspects and embodiments described herein.</figref><figref num="7">FIG. 5 is a flow diagram illustrating an example of an embodiment not intended to be limiting for processing media content collation according to the various aspects and embodiments described herein.</figref><figref num="8">FIG. 5 is a flow diagram illustrating an example of an embodiment not intended to be limiting with respect to a live media stream, according to the various aspects and embodiments described herein.</figref><figref num="9">FIG. 6 is a block diagram showing exemplary computer devices arranged according to the various aspects and embodiments described herein.</figref><figref num="10">It is a block diagram which shows the exemplary network environment by various aspects and embodiments of this disclosure.</figref>

0011Overview Various aspects of the disclosure are described herein with reference to drawings in which similar reference numbers are used in their entirety to represent similar elements. In the following description, for purposes of explanation, many specific details are provided to provide a deeper understanding of one or more aspects. However, it should be understood that certain aspects of the present disclosure may be implemented without these particular details or using other methods, components, components and the like. In other examples, well-known structures and devices are shown in the form of block diagrams to facilitate the description of one or more aspects.

0012According to one or more embodiments described in this disclosure, the user may provide personal information, demographic information, location information, property information, confidential information, etc. in connection with aspects of data collection. It will be understood that you can opt out. In addition, one or more embodiments described herein may provide anonymization of collected, received, or transmitted data.

0013According to one or more embodiments described in the present disclosure, the audio matching system receives an audio signal excerpt (probe) as input and a corresponding audio excerpt in a large repository of reference audio signals. Try to find. For example, a cell phone records music playing in a noisy environment (eg, a noisy bar, or elsewhere), which matches a noisy probe with a large repository of references. Can be utilized by a matching system that can return information about the music being played by. When a matching system receives an audio clip from a client, it can work to minimize delay by performing a real-time matching of an audio sample (music recording) with a reference object (eg, an audio reference). Matching can be performed in real time upon receipt of the recorded audio stream, based on various parameters such as the time and / or increment of the received audio.

0014For example, using a database of tens of millions of audio references, the system can possibly find multiple matching objects that meet a given matching object threshold. The resulting duplication of matching objects can occur for a number of reasons, including duplication in a set of references, or the same track can be present on both the compilation album as well as the main EP album. As another example, the resulting audio reference may come from both the original version and a less popular remix that matches the probe audio sample. Although audio content is described herein as an example, other media content is also included by means such as video content, image content and the like.

0015In the context of a streaming matching system, return a matching thing that is exactly the matching thing available to return and has previously passed a predetermined threshold to determine the matching thing. Or it can be difficult to be sure of waiting for a more desirable matching thing to become available. In various aspects of the disclosure, component real-time ranking systems are less popular in matching systems until they are ranked and output more reliable and consistent objects. Acts to hold back or hold a reference that can be. The receiving component receives the first part of the media streaming content to determine if the matching thing is available in the media content reference. Upon receiving at least one reference audio, the comparison component identifies at least one reference audio sample in real time by comparing the first portion of the audio streaming content with multiple reference audio samples. The matching component identifies, based on comparisons, a good match between the first part of the audio streaming content and the multiple reference audio samples, and the ranking component is among the multiple reference audio samples. Build a first ranking score for the matching reference audio sample.

0016An example that is not intended to hold back results and limit real-time rankings in a streaming matching system. Reference here is to describe System 100. System 100, for example, collates objects represented by feature vectors and ranks the results continuously updated, where the system receives the objects in a state such as a video or audio stream for comparison. At the same time, it is possible to output the resulting matching objects based on the ranking, which is determined at the same time. System 100 may include memory for storing computer-executable components and a processor for executing computer-executable components stored in memory, an example of which can be found by looking at Figure 9. .. In addition, system 100 may include receiving component 102, comparing component 112, matching component 118, and ranking component 124.

0017The receiving component 102 may be configured to receive the first feature vector 104, represented as U. U can be associated with reference object 106. The receiving component 102 may also receive a second feature vector 108, represented by V, which may be associated with the comparison object 110. The first feature vector 104 (eg, U) and the second feature vector 108 (eg, V) can be d-dimensional vectors. Where d is, for example, any positive integer value. The d-dimensional U or V may be associated with the individual features associated with the reference object 106 and the comparison object 110, respectively, as further details are provided in connection with FIG.

0018If we turn our attention to FIG. 2 while still referring to FIG. 1, we see schematic diagram 200. Schematic 200 shows various examples of reference or comparison objects. For example, objects 106, 110 may be associated with at least one of image 202, audio 204, and / or text 206. If objects 106, 110 are associated with image 202, feature vectors 104, 108 (eg, U and V) may include image features, such as pixel values. As an example, the first feature vector 104 may contain pixel values for pixels in the reference image and the second feature vector 108 may contain pixel values for pixels in the comparison image. In other embodiments, the feature vectors 104, 108 may include wavelets or other features (eg, including local features) associated with the reference image and the comparison image. In addition, the feature vector may include an audio fingerprint that has one or more identifiers of the content, or a compact content-based signature that summarizes or identifies the audio content. If objects 106, 110 are associated with audio 204, for example, the first feature vector 104 may include a wavelet value (or other feature associated with the audio) for the wavelet of the reference audio. , The second feature vector 108 may include a wavelet value (or other audio-related feature) for the wavelet of the comparative audio. If objects 106, 110 are associated with text 206, for example, the first feature vector 104 is a word count (or other feature associated with the text) for various words contained in the reference text. The second feature vector 108 may include word counts (or other features associated with the text) for the various words contained in the comparison text.

0019Returning to FIG. 1, when the comparison component 118 receives the first part or segment of the comparison object (eg, audio streaming content), it is among the reference objects (eg, reference audio samples) of one or more volumes. Reference objects 106 (eg, audio samples) can be configured to identify in real time. The comparison component 118 compares a first portion of audio streaming content (eg, a predetermined number of feature vectors, a predetermined time period, and / or a quantity) with a plurality of reference audio samples.

0020The matching component 124 may be configured to identify a well-matched object between the first part of the comparison object 110 and the plurality of reference objects 106, based on the comparison. For example, matching can be identified based on a consistency score of 120 between one or more feature vectors U and V. Therefore, a well-matched thing 126 for U and V is sufficient between the reference object 106 (or part of the reference object 106) and the comparison object 110 (or the relevant part of the comparison object 110). Can mean things that match.

0021In addition or / or the consistency score 120 may be determined based on parameter 122. Parameter 122 may indicate, for example, whether to use an indicator function or distance measure to determine a consistency score 122 to determine a good match. The ranking component 124 may be configured to construct a first set of matching reference results 128, represented as P, and a second set of matching reference results 130, represented as Q. P can be constructed based on an array of the first elements of the U dimension (eg, the first part of the first feature vector 104), and Q can be the second of the U dimensions. It can be constructed on the basis of an array of elements (eg, the first part of the second feature vector 108), and the array of first elements and the array of second elements are the first subset of the comparison object 110 and It may contain each of the second subsets. P and Q can be matching results of adjacent element sequences that differ with respect to each other. The array of first elements or the array of second elements can be a subset of both the other subset and the comparison object 110.

0022FIG. 3 shows a system 300 operating as a collation system according to various embodiments described herein. For example, the system 300 continuously matches a comparison object, such as a sample audio stream, with a reference object (eg, reference audio) to continuously rank the results of matching objects and a predetermined score threshold. It works to generate higher reliability to output the result of the matching object by holding the matching object until it satisfies. For example, the system 300 includes components as described above, as well as a scoring component 302 that compares ranking scores, updates rankings, and determines the results of well-matched items.

0023The scoring component 302 is processed and ranked by the ranking component 124 in real time as the first part of the comparison object 110 is processed and ranked through the receiving component 102, the comparison component 112, the matching component 118, and the ranking component 124. It is configured to receive the generated collation results. The scoring component 302 may receive a first set of rankings 128, and when an additional portion of the comparison object 110 is processed to determine a matching object in the reference object, the scoring component matches. Dynamically update the ranking of the results of the objects that match, and use additional results with a certain level of reliability that meet a given threshold for the well-matched objects 126 to get the results of the matching objects. It can be updated and / or determined to meet a higher second predetermined score threshold for immediate output.

0024In certain embodiments, the system 300 has a first set of rank scores and corresponding matching objects associated with the determined rankings for further processing or matching of additional parts of the comparison object 110. Works to hold the result of (match reference object). The ranking of the first set can be received with the results of matching objects, and the scoring component is ranked to meet a second predetermined threshold score for the ranked matching objects. Any of the results can be released. The second predetermined threshold (score threshold) has a higher reliability threshold level and can differ from the well-matched mono threshold for matched references. Here, well-matched objects can provide a certain level of matching between objects in the comparison, and the score threshold provides a higher likelihood of matching based on parameter 304. A second predetermined threshold (score threshold) is met with additional criteria such as parameter 304 and / or collation results as determined from a further portion of the comparison object calculated in the comparison between reference objects. May contain more quantity of feature vectors in.

0025For example, the scoring component 302 uses an additional part or segment of the audio stream to determine if the ranking of matching objects is sufficient to output as matching objects. Various parameters 304 may be used to determine if the additional collation operations / processes that were required for higher reliability. If the result of the matching object does not meet the second predetermined threshold for immediate output, then the first set of ranked matching objects is retained and the comparison object 110 It can be dynamically updated based on additional iterations of the matching process with additional parts.

0026For example, the scoring component 302 operates to perform a score comparison between the result 128 of the first match reference and the result 130 of the second match reference in the match reference audio sample. The second result 130 may include a second ranking score associated with a second or additional portion of the audio streaming content (eg, comparison object 110). The scoring component 302 performs additional iterations on the same comparison object or media stream based on a larger amount of analyzed streaming content and / or a predetermined amount of media streaming content received by System 300. By changing the ranking based on the result of, it can act to iteratively and dynamically change the score of the ranked matching objects. In addition or / or, the scoring component 302 meets the score threshold by updating the results ranked according to the ranking 130 of the second set from the comparison between the rankings of the first and second sets. Or weighting the rankings received in the first set of rankings 128 to release any retained results that meet a higher degree of score threshold than the other resulting reference matches determined. Can work with.

0027In certain embodiments, the second part of the comparison object 110 (eg, V [val)._d]) Second match reference result 130 includes a first part and a subset of audio samples (eg, to further determine the strength of the result and / or the result of the additional match reference determined. Comparison object) can be included. The ranking component 124 may work with the scoring component 302 to dynamically change the first ranking score associated with the matching reference audio sample based on the score comparison. The result of a matching thing may be released for output in response to reaching a second predetermined threshold based on a quantity of processed comparison objects 110 or at least one of a quantity of time. Matching mono results determined for output can be remembered prior to release based on the associated ranking score and are more reliable about the release results and their ranking accuracy. Can have more. For example, if the results of multiple different matching things can occur in different albums, EP originals, remixes, etc., then the first part of the media stream is compared and the things are well matched. Rather than releasing results immediately when identified as, a given number of results may be released with greater accuracy or reliability for a promising matching object.

0028The scoring component 302 can make continuous comparisons of rank scores from various parts of the comparison object based on parameters or measures. The parameters or rating measures are, for example, whether each matching reference audio sample is individually available for sale, each popularity of the matching reference audio sample, whether each matching reference audio sample is an original album and / or It may contain one or more weightings or scores that are compilation albums and / or that are based on and / or indicate the respective sales volumes of the matching reference audio sample.

0029In addition, the results of the identified matching objects can be ranked according to the intensity of the matching objects, and the scoring component further determines the score of the matching objects based on the parameters. Can be done. For example, the ranking of the first set may be received from the first part of the media streaming content that is received and the results of multiple matching references such as References A, B, and C are found. Each of these references can be ranked based on the strength of the values between the feature vectors between some comparison content / objects and the reference content / objects. The reference can be ranked and communicated with the scoring component 302 before being output as a matching reference. The scoring component 302 then holds the matching object until the score is determined, the other additional parts of the comparison object are processed, and / or the amount of time for the matching operation of the matching component elapses. obtain. For example, only one or more (eg, A and B) may be output as matching objects. In this case, C has not passed the predetermined threshold.

0030The scoring component 302 may act to hold each of the matching objects and further analyze it based on the parameter / evaluation scale 304. In some situations, one or more matching objects may meet a second predetermined threshold based on a set of parameters. The scoring component 302 may then output these matching objects from the system 300 as a result and / or continue to retain the results of the matching objects. The scoring component 302 can continue to receive matching objects and associated rankings and compare them with the results they hold. For example, continue to receive reference A as a matching thing, however, with respect to other references B and C, additional data to match that reference (an additional amount of received audio stream, or an addition of comparison objects). Can be ranked higher based on (part of). Therefore, the scoring component 302 determines whether reference A meets a second predetermined threshold and to what extent reference A meets or does not meet the ranks in the score. An increase can occur.

0031See Figure 4 to show an exemplary scoring component of a matching system that produces a score or merit scale to output a match reference. The scoring component 302 can generate a score 400 that is specific to the music content, as represented by the comparison object 110 as described above, and / or other types of content (eg video, TV). Appropriate scores can be defined for the program). As an example, the score 400 is represented as an integer value with one or more bits and can contain many components. It is also assumed that other integer values with different bit lengths may have one or more components arranged in various orders and / or have different meanings in their representation.

0032For example, the high-order bit 402 can be used to track availability for sale. If the high-order bit 402 is one (1), then object 110 (eg, songtrack, for example) is individually available for sale, otherwise it is zero (0). In addition, the following 4-bit set 404 can be used as a popular rating measure. This rating scale is provided by heuristics such as web searches and / or hits for an object or audio track (eg views, downloads, etc.) and can be specified as a binary value represented by 4 bits or more. In another embodiment, another bit (eg, the sixth bit) may represent a matching reference as an original EP or compilation album in the case of an audio reference. For example, 1 indicates that the track is an original EP, and can be 0 if it is a compilation album, and vice versa. The remaining bits 408 can be used to encode sales volume data and / or other data related to potential match content references. As a result, the higher the score, the more favorable the matching system will return, and when it meets a second predetermined threshold for release as a result of matching objects. The degree or reliability level that the reference has is higher. The order and meaning of each component of the score (corresponding parameters or evaluation scales) can vary and is not limited to the examples described above to illustrate the score 400. For example, the score may be expressed as an integer value that can contain any number of components. Each of these components may have a group or set (s) of bits occupying the most significant to least significant bit positions, based on the relevance importance of the feature.

0033The scoring component 302 basically uses parameter 304 (evaluation scale criterion) in the priority order to determine which of the matching references should be retained before release or output. , Two scores can be compared. Therefore, a score of 400 is for reference matches (eg, audio tracks with a first set of matching references and / or an audio track with a second set of identical references with different numbers of matching feature vectors, and / or different references). Compare any pair, recalculate the ranking for the list of matching things, and / or rank the score to reset the ranking for the remembered matching things that will be output. Provides a means of communicating with the attachment component 124.

0034In another embodiment, a predetermined score threshold T (a second predetermined threshold or score threshold) is defined by the scoring component 302, above which the scoring component releases as a matching thing. Or do not hold back the matching thing after outputting. For example, the threshold is T = It can be a score of 400, shown as 11111000000000000000000000000000. Any matching object above this threshold is considered sufficient to be immediately returned to the user. Matching objects below the threshold are matched until the system as described above receives and matches a probe (eg, comparison object, audio sample stream, etc.) and / or an additional portion (quantity) of the probe for at least N seconds. Cannot be held back. For example, a value of N = 5 seconds can be used. In addition or / or the quantity of the probe or comparison object is also a match generated from the first part (eg, one or more features) until an additional part is also matched with or separately from the first part. It can be specified so that it can hold the object that is being used. Either strengthening, weakening, and / or abandoning those that do not meet additional thresholds (eg, T) by retaining the initial matching objects. Higher reliability can be achieved by doing so.

0035In most cases, the client (eg, the client device) will see the first matching thing, but for completeness, the systems mentioned above are ranked matching. It works to return a complete list of things in real time. At the time of matching, the system may receive probe clips (eg, comparison objects) from the client and initiate clipping and matching of the clips. Depending on the matching component's identification of matching objects or sets of matching objects, the ranking component ranks them and immediately returns and removes those above the threshold T. Here the system waits to release the result until it receives a more matching thing. Each time a new set of matching things is received, they are merged, ranked, and seconded into the remaining, non-returned / holding set of matching things. Can be compared against a given threshold or score threshold. After receiving and matching data for N seconds and / or N cycle intervals, the system may read the ranked set of all matching things found so far and return them. Any subsequent matching objects after this point may be ranked and immediately returned to the user.

0036Here, with reference to FIG. 5, the ranking component 124 according to the various aspects described herein is shown. The ranking component 124 may include, for example, a scoring component 302 that acts to determine the score of a matching reference that meets the threshold of well-matched objects. The ranking component 124 may further include a retention component 502 and a merge component 504.

0037Retention component 502 retains a matching reference audio sample that has a well-matched object or meets the matching object threshold. The result of any matching reference that meets the score threshold based on parameter 304 can be released immediately as one or more released matching mono 506. The retention component 502 outputs until at least a second set of match reference audio samples is determined from the additional part of the comparison object (probe sample) and then the initial match reference results and the first associated with them. It can hold a ranking score. Therefore, the ranking component 124 continuously compares the rank scores of the matching reference results and meets the score threshold for which result is released as one or more released matching mono 506. Can act to determine.

0038Retention component 502 may retain the result of a match reference based on a predetermined parameter that includes at least one of the time of the comparison process or the amount of audio streaming content compared. In addition, the parameters are whether each match reference audio sample is individually available for sale, each popularity of the match reference audio sample, whether each match reference audio sample is an original album or a compilation album. Alternatively, it may include indicators for the respective sales volumes of the matching reference audio samples, each of which determines the rank score, and each of the matching references meets a score threshold (a second predetermined threshold as described above). Can be weighted differently with respect to whether or not.

0039After releasing the best or best match reference results based on the score ranking after processing a certain amount of time and / or a certain amount of comparison objects, the retention component 502 is released Matching Mono 506. It may act to release all matching objects as, or as a subset of matching objects that are determined to meet the score threshold to a higher degree than other reference matches.

0040The merge component 504 merges the match reference audio sample with at least one additional match reference audio sample that has a second ranking score of things that closely matches the second part of the audio streaming content. Can work further. For example, when processing additional parts or segments of an audio probe or comparison object, additional match references may be found and included from the first set of matching objects held. Or or in addition, the merge component 504 is new, such as when the additional amount of audio probe or comparison object data may indicate a stronger matching thing and / or an increase in one or more parameters 304. It can act to merge score ranks by updating the ranks with the results of the ranked match references.

0041An example that is not intended to limit the method for holdback and real-time ranking of results in a streaming matching system. Figures 6-8 show various methods depending on the subject of the disclosure. Although the methods are presented and disclosed as a series of processes for brevity, the subject matter of disclosure is not limited to the order of the processes, and some processes are shown and described herein in a different order and / or. It may occur at the same time as other processing. Alternatively, one of ordinary skill in the art will understand and understand that the method can be interpreted as, for example, a series of interrupted states or events in a phase diagram or the like. Moreover, not all of the illustrated processes need to implement the method according to the subject of disclosure. In addition, it will be appreciated that the methods disclosed in this disclosure can be stored in the product to facilitate transport and transmission of such methods to a computer or other computer device.

0042Here, with reference to FIG. 6, an exemplary method 600 for a matching system according to the various disclosed embodiments is shown. At reference number 602, a probe audio sample is received. The probe audio sample may include, for example, any comparison object of media content (eg, audio, video, textual data, etc.) that may include a stream of media content, as described above with respect to FIG.

0043At reference numeral 604, method 600 comprises comparing the probe audio sample with a plurality of reference audio samples in order to identify at least one matching reference audio sample. For example, comparing a probe audio sample to multiple reference audio samples to identify one or more matching reference audio samples is at least matching the generation of the probe audio sample thresholds. It may include identifying the respective fingerprints of multiple reference audio samples that match the fingerprints of the probe audio sample by determining the threshold of or a well-matched thing.

0044At reference number 606, each of the ranking scores is assigned to a matching reference audio sample as it identifies a plurality of matching reference audio samples. For example, assigning each of the ranking scores to a matching reference audio sample may include generating each of the ranking scores based on one or more rating scales / parameters. The parameters are an availability rating scale that indicates whether each matching reference audio sample is individually available for sale, a popularity rating scale that indicates the popularity of each matching reference audio sample, and each matching reference audio sample. It may include a compilation rating scale indicating whether is an original album or a compilation album, and / or a sales rating scale indicating the volume of each of the matching reference audio samples.

0045For example, the ranking score contains a specific bit length, such as a 32-bit vector, or some other bit length that indicates various parameter criteria, thereby determining whether to release matching objects as output results. obtain. As an example, the high-order bits can represent a measure of availability (eg, whether or not they are publicly available for sale), and the other bits can be used for web searches and / for match reference results. Or it can represent a popular rating scale (eg, heuristic data) such as the number of views or downloads, and the other bit is a compilation rating scale or other metadata (eg, originally an original EP or compilation album). If), a bit about the number of sales data, and / or other criteria used to generate a factor ranking score may be indicated.

0046In certain embodiments, the resulting matching objects (eg, matching reference audio samples) can be ranked based on the degree of correlation between fingerprints or orders, and ranked matching. An item can be scored based on a set of parameters / evaluation scales as described above.

0047In another embodiment, the method flow assigns each of the ranking scores and determines and / or has already been assigned a matching one with a higher ranking than the one initially identified among the plurality of matching reference audio samples. After correcting the existing rank score, it can be continued by continuously comparing the probe audio sample with multiple reference audio samples. For example, additional parts of the probe sample may be further matched to generate additional matching references and / or identify additional features / dimensions in the comparison for matching objects. A first set of matching objects in the initial volume of the sample can identify a large number of matching objects, while an additional volume of probe sample finds additional matching objects. At the same time, it can strengthen or weaken the already determined matching thing. For example, when additional matching objects are found, and / or the occurrence of features where the matching object ranks against the matching object threshold is less than was done in the previous part of the probe sample. When pulled down from, the parameters can be weakened or pulled down against the matching thing they are holding. The parameters may be applied less or more, for example, to the results of the matching objects held. Here, the additional result reduces the strength of the matching thing (eg, audio remix or other modified recording, and / or similar media). Therefore, by retaining the initial results from the output, one or more matching references with the highest score ranks can be output for further comparative analysis and are more reliable than they would otherwise be. Have a level.

0048Here, with reference to FIG. 7, an exemplary method 700 for a matching system, according to the various aspects described in the present disclosure, is illustrated. At reference numeral 702, method 700 includes receiving a first portion of the probe audio sample via a processor. For example, the first part is the amount of probe sample that is sufficient for a certain amount of time to receive the probe sample and / or the matching process (eg, media content, one second, number of feature vectors detected, etc.) It can be determined from 1/50).

0049At 704, the first portion of the probe sample is compared to multiple reference audio samples to identify multiple matching reference audio samples. In 706, identify one or more matching reference audio samples that meet the well-matched mono threshold.

0050At 708, a matching reference audio sample that meets the well-matched mono threshold is determined based on a set of parameters.

0051At 710, it outputs and then holds a matching reference audio sample. In some embodiments, the ranking or score threshold is set, while retaining others, prior to the release and / or additional collation of probe samples, etc. with respect to the additional data that will be received. Such references with their respective assigned ranking scores that are exceeded need not be set and retained for immediate release. In another embodiment, the matching reference sample may be held for a period of time and then set to release a predetermined number of matching reference results (eg, matching audio sample results). In addition, or after a predetermined amount of time to output a matching reference audio sample with each of the ranking scores exceeding the ranking threshold, the method has held multiple with each of the assigned ranking scores. Guaranteed to release for the first output of the best matching thing, which can be continued by outputting the matching reference audio sample, or without further holding the found matching thing. obtain.

0052Processing at 708 may be continued further with reference to Method 800 in FIG. In 802, a matching reference audio sample that exceeds or meets the score threshold, determined based on parameters as described above, may be retained (eg, by a data store or other memory) before being released. In 804, the No. 1 of the probe audio samples for multiple reference audio samples, in order to identify any additional matching reference audio samples and / or to enhance / weaken the results of the matching references held. Can be done for matching with part 2. Ranking scores are continuously based on further comparisons and the set of parameters for the matching reference audio sample held, and with any additional matching reference audio samples identified from the second part of the probe sample. Can be released to or repeatedly. In 806, based on the re-evaluated ranking score, it may be released after retaining matching objects and / or for immediate release of additional matching results. For example, release as output a subset of match reference audio samples that meet a higher score threshold than the other match reference audio samples in the set (a set containing the matching results held and additional match reference audio samples). Can be done. And it is not always necessary to retain the results of additional matching objects so that additional parts of the probe sample that produce the matching objects can be released immediately as output without any retention or holdback. .. Alternatively, further results may be retained and continued to be reassessed by comparison with any additional matching objects. Match reference audio sample and additional match reference audio sample.

0053Illustrative network and distributed environments Any of the various embodiments described herein can be implemented as part of a computer network or in connection with any computer or other client or server device that can be deployed in a distributed computing environment and can include media. Those skilled in the art will appreciate that they can connect to different types of data stores. In this regard, the various embodiments described herein are any computer system or environment with any number of memory or storage units, and any number of applications that occur across any number of storage units. And can be implemented in the process. This includes, but is not limited to, a network environment with remote or local storage or an environment with server and client computers deployed in a distributed computing environment.

0054Distributed computing provides the sharing of computer resources and services through the transmission and exchange of computer devices and systems. These resources and services include information exchange, cache storage, and disk storage for objects such as files. These resources and services also include sharing of processing power across multiple processing units for load balancing, source expansion, processing specialization, and so on. Distributed computing leverages network connectivity, which allows clients to harness their collective capabilities for the benefit of the entire enterprise. In this regard, the various devices may have applications, objects or resources that may be involved in the mechanisms described with respect to the various embodiments of the present disclosure.

0055FIG. 9 provides a schematic diagram of an exemplary network or distributed computing environment. A distributed computing environment can include programs, methods, data stores, programmable logic, etc., as represented by applications 930, 932, 934, 936, 938, computing objects 910, 912, etc. and computing objects or Equipped with devices 920, 922, 924, 926, 928, etc. Computing objects 910, 912, etc. and computing objects or devices 920, 922, 924, 926, 928, etc. are personal digital assistants (PDAs), audio / video devices, mobile phones, MP3 players, personal computers, tablets, laptops, etc. It can be understood that it can include different devices such as.

0056Each of the computing objects 910, 912, etc. and the computing objects or devices 920, 922, 924, 926, 928, etc., either directly or indirectly, via the communication network 940, one or more. It can communicate with other computing objects such as 910, 912 and computing objects or devices such as 920, 922, 924, 926, 928. Although shown as a single element in FIG. 9, the network 940 may include other computing objects and computer devices that provide services to the system of FIG. 9 and / or a plurality of mutuals. The connected network is not shown but may be represented. Each of the compute objects 910, 912, etc. or the compute object or device 920, 922, 924, 926, 928, etc. may also utilize the API Applications such as 930, 932, 934, 936, 938. , Or other objects, software, firmware and / or hardware suitable for communication or implementation with various embodiments of the present disclosure.

0057There are various systems, components, and network configurations that support a distributed computing environment. For example, computing systems can be connected together by local or widely distributed networks, by wired or wireless systems. Many networks are now connected to the Internet, which provides infrastructure for widely distributed computing and includes many different networks, as any network infrastructure is described in various embodiments. It can be used for exemplary communications originating from various systems.

0058Therefore, a host of network topology and network infrastructure such as client / server, peer-to-peer, or hybrid mechanism can be used. A "client" is a member of a class or group that uses the services of another unrelated class or group. A client is a process that requests a service provided by another program or process, and can be, for example, roughly a set of instructions or tasks. The client may be a process that utilizes or uses the requested service without having to "know" any operational details about the other program or service itself.

0059In a client / server mechanism, especially a network system, a client is usually a computer that accesses another computer, eg, a shared network resource provided by a server. In the explanatory diagram of FIG. 9, as an example not intended to be limiting, a computing object or device 920, 922, 924, 926, 928, etc. can be considered as a client, and a computing object 910, 912, etc. Can be thought of as a server. Computing objects 910, 912, etc. receive data from client computing objects or devices 920, 922, 924, 926, 928, etc., store data, process data, client computing objects or devices 920, 922, 924, etc. Provides data services such as sending data to, 926, 928, etc., but any computer can be considered a client, server, or both, depending on the circumstances.

0060Servers are generally accessible to remote computer systems via remote or local networks such as the Internet or wireless network infrastructure. Client processes may be active in the first computer system, and server processes may be active in the second computer system and communicate with each other over the communication medium and are therefore distributed. It provides functionality that allows multiple clients to take advantage of the server's information gathering capabilities.

0061In a network environment where the communication network / bus 940 is the Internet, for example, computing objects 910, 912, etc., client computing objects or devices 920, 922, 924, 926, 928, etc. are hypertext transfer protocols (HTTP), etc. It can be a web server that communicates via any of a number of known protocols. Objects 910, 912, etc. may function as client computing objects, or devices 920, 922, 924, 926, 928, etc. may function as distributed computing environments.

0062Illustrative computer device As mentioned above, advantageously, the techniques described herein may be applicable to any device suitable for carrying out the various embodiments described herein. Handheld, portable and other computer devices, and all types, for use related to various embodiments, for example, wherever a device wants to read or write transactions from or to a data store. Compute object is conceivable. Therefore, the following general-purpose remote computer is shown below in FIG. 10, but it is only an example of a computer device.

0063Although not required, embodiments are partially implemented through an operating system for use by developers of services relating to devices or objects, and / or various embodiments described herein. It may be included in application software that operates to perform one or more functional aspects. The software may be described in the general context of computer executable instructions such as program modules executed by one or more computers such as client workstations, servers, or other devices. Those skilled in the art will appreciate that computer systems have various configurations and protocols that can be used to communicate data, and therefore do not limit themselves to specific configurations or protocols.

0064As such, FIG. 10 shows an example of a suitable computing system environment 1000 that may implement one or more embodiments of the embodiments described herein, but as clarified above, computing. The Ing System Environment 1000 is only an example of a suitable computing environment and is not intended to make any limited suggestions regarding the range of use or functionality. The computing environment 1000 should not be construed as having any dependency or requirement with respect to any one or combination of components shown in the exemplary operating environment 1000.

0065Referring to FIG. 10, an exemplary remote device that implements one or more embodiments comprises a general purpose computer device in the form of computer 1010. The components of the computer 1010 may include, but are not limited to, a processing unit 1020, a system memory 1030, and a system bus 1022 that connects various system components, including system memory, to the processing unit 1020.

0066The computer 1010 comprises various computer-readable media and can be any available medium accessible by the computer 1010. System memory 1030 may include computer storage media in the form of volatile and / or non-volatile memory such as read-only memory (ROM) and / or random access memory (RAM). As an example, but not limited to, memory 1030 also includes an operating system, application programs, other program modules, and program data.

0067The user may enter commands and information into the computer 1010 via the input device 1040. Also, a monitor or other type of display device is connected to system bus 1022 via an interface such as the output interface 1050. In addition to the monitor, the computer also has other peripheral output devices such as speakers and printers that can be connected via the output interface 1050.

0068Computer 1010 may operate in a networked or distributed environment with a logical connection to one or more other remote computers, such as remote computer 1070. The remote computer 1070 may be a personal computer, server, router, network PC, peer device or other common network node, or any other remote media consuming or transmitting device, or in connection with computer 1010. It may include any or all of the elements described above. The logical connection shown in FIG. 10 comprises network 1072, such as a local area network (LAN) or wide area network (WAN), but may include other networks / buses. Such network environments are commonplace in homes, offices, enterprise wide computer networks, intranets, and the Internet.

0069As mentioned above, exemplary embodiments have been described in the context of various computer devices and network mechanisms, but the underlying concept is any network system in which it is desirable to expose or consume the medium in a flexible manner. And may be applied to any computer device or system.

0070The word "exemplary" is used herein to mean an alternative to an example, example, or figure. For the avoidance of doubt, the matters described herein are not limited to such examples. In addition, any aspect or design described herein as "exemplary" need not necessarily be construed as preferred or desirable with respect to other aspects or designs and is known to those of skill in the art. Nor does it mean excluding equivalents, exemplary structures, and techniques. In addition, to avoid misunderstanding, as long as terms such as "prepare," "have," "include," and other similar words are used in either the detailed description or the claims. Such terms are intended to be inclusive, as is the term "contains", as an open diversion without excluding any additional or other elements.

0071Computer devices generally include a variety of media, including computer-readable storage media. The computer-readable storage medium can be any available storage medium that is accessed by a computer and has non-temporary properties and can include both volatile and non-volatile media, removable and non-removable media. By way of example, but not limited to, computer-readable storage media are implemented in connection with any method or technique for storing information such as computer-readable instructions, program modules, structured data, or unstructured data. Can be done. Computer-readable storage media can be used to store desired information, such as RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic. It may include, but is not limited to, cassettes, magnetic tapes, magnetic disk storage, or other magnetic storage devices, or other tangible and / or non-temporary media. Computer-readable storage media are accessed by one or more local or remote computer devices for various processing of information stored by the medium, for example, via access requests, queries, or other data retrieval protocols. obtain.

0072As mentioned above, the various techniques described herein may be implemented in connection with hardware or software, or in combination of both, if desired. Terms such as "component" and "system" as used herein are likewise computers associated with any entity of hardware, hardware and software combination, software, or software at runtime. Is intended to be referred to. For example, components may be, but are not limited to, processes, processors, objects, executables, threads for execution, programs, and / or computers running on the processor. As an example, both the application running on the computer and the computer can be components. One or more components may exist within a process and / or thread for execution, and components may be localized to one computer and / or distributed among two or more computers. Good. In addition, the components are stored on dedicated hardware, general-purpose hardware specialized by the execution of software that allows the hardware to perform certain functions (eg, encoding and / or decoding), computer-readable media. It may work in the form of software, or a combination thereof.

0073We have described the system described above for interactions between several components. It is understandable that such systems and components may include components or specific subcomponents, parts of specific components or subcomponents, and / or additional components that comply with the various substitutions and combinations described above. Also, sub-components can be implemented as components that are communicated with other components rather than being contained in the parent component (hierarchically). In addition, one or more components may be combined into a single component that provides aggregated functionality or split into several separate subcomponents, or any one, such as the management layer. Also note that multiple middle layers may be provided to communicate and connect with such subcomponents to provide integrated functionality. Any component described herein may also interact with one or more other components not specifically described herein, but generally known to those of skill in the art.

0074Given the exemplary system as described above, the methods that can be implemented by the subject matter described will be more preferred by reference to the flowcharts of the various figures. For the sake of brevity, the method is shown and described as a series of blocks, but the subject matter of the allegations is that some blocks are described and / or described herein in a different order. It may occur at the same time and is not limited to the order of blocks. It is understandable that the flow chart shows a discontinuous or branch flow and that the order of various other branches, flow paths, and blocks can be implemented to achieve the same or similar results. Moreover, the illustrated blocks are not necessarily required to implement the methods described below.

0075In addition to the various embodiments described herein, other similar embodiments have been used or modified and added in order to perform the same or equivalent function of the corresponding embodiment without deviation. It will be appreciated that it can be done in the described embodiments. In addition, multiple processing chips or devices may share the performance of one or more of the features described herein, and similarly, storage may be provided across multiple devices. Therefore, the invention is not limited to any one of the single embodiments, but rather may be construed as being included in the spirit and scope of the appended claims.

0076100 systems 102 Receiving component 104 First feature vector 106 Reference object 108 Second feature vector 110 Comparison object 112 Comparison component 118 Matching component 120 consistency score 122 parameters 124 Ranking component 126 Things that match well 128 Results of the first set of match references 130 Second set match reference results