JP5453937B2 - Genetic processing apparatus, genetic processing method, and genetic processing program - Google Patents

Description

  The present invention relates to a genetic processing apparatus, a genetic processing method, and a genetic processing program.

  A method for generating a filter or a converter using evolutionary computation such as a genetic algorithm or genetic programming is known (see Non-Patent Document 1). According to the method of generating a filter or a converter using such evolutionary computation, fewer filters or converters having a complicated structure that are optimal for each case and difficult to obtain analytically are obtained. It can be designed with effort and time.

Masayoshi Maebuchi and two others, “Study on Image Filter Design Using Genetic Algorithm” [online], Computer Utilization Education Council, [March 20, 2008 search], Internet <URL: http: //www.ciec. or.jp/event/2003/papers/pdf/E00086.pdf>

  By the way, the method of generating a converter by such evolutionary calculation requires a huge number of generations until a target converter is obtained. Therefore, such a method has a huge calculation cost until the target converter is obtained. In addition, in the process of evolutionary computation, there may be cases where significant progress or change is obtained, or there is a case where the progress is stagnated, but it is difficult to specify the cause.

  Therefore, an object of the present invention is to provide a genetic processing apparatus, a genetic processing method, and a genetic processing program that can solve the above-described problems. This object is achieved by a combination of features described in the independent claims. The dependent claims define further advantageous specific examples of the present invention.

  In order to solve the above-described problem, in the first aspect of the present invention, genetic data is generated from at least one previous generation converter including a plurality of processing components that process input data and output processing results as output data. A generation unit that generates a current generation converter by processing, a fitness calculation unit that calculates a fitness for conversion from learning input data to learning target data for each of the current generation converters, and between generations A genetic processing device, and a genetic processing method for the genetic processing device and a genetic processing. Provide a program.

  It should be noted that the above summary of the invention does not enumerate all the necessary features of the present invention. In addition, a sub-combination of these feature groups can also be an invention.

1 shows a configuration of a genetic processing apparatus 10 according to the present embodiment. An example of the converter 20 which combined the processing component 22 which concerns on this embodiment in series is shown. An example of the converter 20 which combined the processing component 22 which concerns on this embodiment with the tree structure is shown. An example of a genetic operation performed on the converter 20 in which the processing components 22 according to this embodiment are combined in series is shown. An example of crossover operation performed with respect to the converter 20 which combined the process component 22 which concerns on this embodiment with the tree structure is shown. An example of the mutation operation performed with respect to the converter 20 which combined the processing component 22 which concerns on this embodiment with the tree structure is shown. An example of the operation | movement flow of the genetic processing apparatus 10 which concerns on this embodiment is shown. An example of the transition of the fitness in each generation which concerns on this embodiment is shown. An example of the processing flow in step S21 of FIG. 7 is shown. An example in which the collection unit 49 according to the present embodiment records a moving image is shown. An example of the processing flow of the weight generation unit 43 in step S22 of FIG. 7 is shown. An example of the slide bar for setting the introduction probability of a plurality of processing components that can be incorporated in the new converter 20 according to the present embodiment is shown. 2 shows an exemplary hardware configuration of a computer 1900 according to the present embodiment.

  Hereinafter, the present invention will be described through embodiments of the invention, but the following embodiments do not limit the invention according to the claims. In addition, not all the combinations of features described in the embodiments are essential for the solving means of the invention.

  FIG. 1 shows a configuration of a genetic processing apparatus 10 according to the present embodiment. The genetic processing apparatus 10 evolves a group of transducers including at least one transducer 20 over a plurality of generations based on evolutionary computation. In this embodiment, each converter 20 has a structure in which inputs and outputs of a plurality of processing components 22 that process input data and output processing results as output data are combined. Then, the genetic processing device 10 generates a converter 20 suitable for converting learning input data into learning target data. The genetic processing device 10 is realized by a computer as an example. Moreover, the converter 20 may be an image filter as an example.

  The genetic processing apparatus 10 includes a converter storage unit 34, a learning input data storage unit 36, a conversion processing unit 38, a learning output data storage unit 40, a learning target data storage unit 42, and a weight generation unit. 43, a fitness calculation unit 44, a selection unit 46, an update unit 48, a collection unit 49, a generation unit 50, a generation information storage unit 51, a display unit 52, and a re-execution unit 53. The converter storage unit 34 stores a plurality of converters 20 having different configurations.

  The learning input data storage unit 36 stores learning input data to be converted by the converter 20. As an example, the learning input data may be a sample of data that is actually given to the converter 20 in an application to which the converter 20 generated by the genetic processing device 10 is applied. When the converter 20 is an image filter, the learning input data may be an image generated or photographed in advance by the user as an example.

  The conversion processing unit 38 sequentially acquires the plurality of converters 20 stored in the converter storage unit 34. The conversion processing unit 38 converts the learning input data stored in the learning input data storage unit 36 by each acquired converter 20 to generate each of the learning output data.

  The learning output data storage unit 40 stores the learning output data generated by the conversion processing unit 38. As an example, the learning output data storage unit 40 stores each of the generated learning output data in association with the converted converter 20.

  The learning target data storage unit 42 stores learning target data that is a target of output data generated by converting the learning input data by the converter 20. The learning target data is, for example, a sample of data that the converter 20 that has input the learning input data should actually output in an application to which the converter 20 generated by the genetic processing device 10 is applied. It's okay. When the converter 20 is an image filter, the learning target data may be an image generated or prepared in advance by a user, for example.

  The weight generation unit 43 updates the weight data representing the weight for each data area in the comparison between the learning output data and the learning target data for each generation. For example, the weight generation unit 43 may generate weight data in which data is arranged in the same manner as learning output data or learning target data.

  Here, the weight generation unit 43 assigns a weight to data having a larger difference for each data between at least one set of the output data for learning and the learning target data, and a weight for data having a smaller difference. Generate larger weight data. For example, the weight generation unit 43 adds a predetermined value to the weight corresponding to data having a difference greater than the threshold in the current generation weight data, and sets a value smaller than the predetermined value to the weight corresponding to data having the difference equal to or less than the threshold. The next generation weight data may be generated by addition. For example, the weight generation unit 43 adds a predetermined value to the weight corresponding to data having a difference greater than the threshold in the current generation weight data, and subtracts the predetermined value from the weight corresponding to data having the difference equal to or less than the threshold. Then, next generation weight data may be generated.

  The weight generation unit 43 stores weight data. Further, the weight generation unit 43 updates the stored weight data, for example, for each generation so that the weight of data having a larger difference between the selected data is larger than the weight of data having a smaller difference.

  The goodness-of-fit calculation unit 44 calculates the goodness of the conversion from the learning input data to the learning target data for each of the current generation converters 20 stored in the converter storage unit 34. Here, the fitness is an index value indicating whether or not the learning input data is suitable for conversion into learning target data. The higher the value, the more the learning input data is converted into learning target data. It is suitable for Further, as an example, the fitness calculation unit 44 calculates the fitness for conversion from the learning input data to the learning target data for each current generation converter 20 by weighting with the weight data.

  The selection unit 46 selects at least one converter 20 among the plurality of converters 20 stored in the converter storage unit 34. In this case, the selection unit 46 preferentially selects the converter 20 having a higher fitness. More specifically, the selection unit 46 selects at least one converter 20 to be left by a technique that models natural selection of living things. As an example, the selection unit 46 selects at least one converter 20 by genetic calculation such as elite selection and roulette selection based on the fitness of each of the plurality of converters 20 stored in the converter storage unit 34.

  The update unit 48 leaves the converter 20 selected by the selection unit 46 among the plurality of converters 20 stored in the converter storage unit 34, and tricks the converter 20 not selected by the selection unit 46. . For example, the update unit 48 hesitates by deleting the converter 20 from the converter storage unit 34.

  The collection unit 49 collects information about the corresponding generation when the increase in the fitness between generations is larger than the reference value. Here, the collection unit 49 may use the highest value of the suitability or the average value of the suitability of each converter 20 in each generation as the fit of the generation. In addition, when the increase in fitness between generations is larger than the reference value, the collection unit 49 collects information on more corresponding generations compared to the case where the increase in fitness between generations is less than or equal to the reference value. May be collected. Here, as an example, the collection unit 49 sets the immediately preceding generation and / or the immediately following generation as the generation to be collected among the generations in which the increment of the fitness is greater than the reference value. Moreover, the collection part 49 may collect the information regarding a corresponding generation, when the increment of the adaptability between generations is relatively large compared with another generation. Further, the collection unit 49 may collect information on the corresponding generation according to the fitness level when the increase in fitness level between generations is larger.

  For example, the collection unit 49 collects at least one of the structure of at least one converter 20 and the learning output data obtained by converting the learning input data by the converter 20. More specifically, the collection unit 49, for example, for each of the generated new converters 20, the type of genetic operation, information for specifying the parent converter 20, and crossover or mutation Collect at least one piece of information for specifying the set of processing components 22 replaced by the above operation. When the converter 20 is an image filter, the collection unit 49 may collect at least one of an output image and a weighted image output from the image filter.

  Further, the collecting unit 49 collects at least one of the structure of the converter 20 having a fitness equal to or higher than the predetermined value selected by the selecting unit 46 and the learning output data obtained by converting the learning input data by the converter. Also good. Here, the collection unit 49 may collect information about more converters 20 when the increment of the fitness between generations is larger than the reference value, compared to a case where the increase is smaller than the reference value. In this case, the collection unit 49 may preferentially select the number of converters 20 corresponding to the reference value from the converters 20 having a high fitness. Furthermore, as an example, the collection unit 49 collects weight data when the increase in the fitness between generations is larger than the reference value. Further, as an example, the collection unit 49 may collect weight data when the increase in fitness between generations is relatively large compared to other generations.

  Further, as an example, the collection unit 49 stores all the converters 20 in a reproducible manner when the increment of the fitness between generations is larger than the reference value. Here, the collection unit 49 may store all the converter groups including at least one converter 20 of the target generation. Further, the collection unit 49 may store structural information of the converter 20 such as a combination of the processing components 22. Further, as an example, the collection unit 49 may store all the converters 20 in a reproducible manner when the increase in fitness between generations is relatively large compared to other generations.

  The generation unit 50 processes genetic data from at least one previous generation including a plurality of processing components 22 that output processing results as output data and at least one converter 20 stored in the converter storage unit 34. The current generation converter 20 is generated by the process.

  The generation unit 50 acquires from the converter storage unit 34 at least one converter 20 remaining in the update process by the update unit 48. Subsequently, the generating unit 50 performs a genetic operation such as crossover and mutation on the acquired at least one converter 20 to generate a new converter 20. Then, the generation unit 50 writes the generated new converter 20 in the converter storage unit 34. Thereby, the converter storage unit 34 can store at least one converter 20 remaining in the update process and a new converter 20 generated by the generation unit 50.

  The generation information storage unit 51 stores information about each generation collected by the collection unit 49. In addition, the generation information storage unit 51 also stores, for example, moving images in which the display unit 52 displays output results in the order of generation recorded by the collection unit 49.

  When one generation is specified during the output of the moving image acquired from the generation information storage unit 51, the display unit 52 displays information on the specified one generation. The display unit 52 may display information on the specified generation and information on several generations before and after the specified generation as information to be displayed. Here, the display unit 52 may display a part of information related to the generations before and after the specified generation.

  The re-execution unit 53 reproduces all the converters 20 of the specified generation in response to an external designation such as a user designation or a change by a random number, etc. Set a processing condition different from the processing condition that generated, and re-execute the genetic processing. For example, the re-execution unit 53 sets at least one of the mutation rate, the crossover rate, and the introduction probability at which each of the plurality of processing components 22 is introduced into the converter 20 as the processing condition. Thereby, the production | generation part 50 can produce | generate a converter on new conditions in the middle of a process.

  Such a genetic processing device 10 includes a conversion process by the conversion processing unit 38, a fitness calculation process by the fitness calculation unit 44, a converter 20 selection process by the selection unit 46, an update process and a generation unit by the update unit 48. The process of generating a new converter 20 by 50 is repeated a plurality of times (for example, a plurality of generations). Thereby, the genetic processing apparatus 10 can generate the converter 20 suitable for converting the learning input data into the learning target data using the evolutionary calculation.

  FIG. 2 shows an example of the converter 20 in which the processing components 22 according to this embodiment are combined in series. FIG. 3 shows an example of the converter 20 in which the processing component 22 according to this embodiment is combined in a tree structure.

  The converter 20 may have a configuration in which a plurality of processing components 22 are combined in series as shown in FIG. Further, the converter 20 may have a configuration in which a plurality of processing components 22 are combined in a tree structure as shown in FIG. Further, the converter 20 may have a configuration having a plurality of output terminals with respect to one input terminal. Further, the converter 20 may have a configuration having a plurality of input ends and a plurality of output ends.

  The converter 20 performs arithmetic processing on the received input data and outputs output data. As an example, the converter 20 is executed by a program that performs operations on input data. Further, the converter 20 may be an arithmetic expression representing the content of the operation to be performed on the input data.

  The converter 20 having a configuration in which the processing components 22 are combined in a tree structure receives input data from the processing components 22 at the end of the tree structure, and outputs output data from the highest processing component 22 in the tree structure. Further, in such a converter 20, the same input data is given to each of the plurality of terminal processing components 22. Alternatively, such a converter 20 may be provided with different input data for each of the plurality of terminal processing components 22.

  Each converter 20 may be an image filter in which a plurality of filter components that respectively convert an input image into an output image are combined. In this case, each processing component 22 receives the image data output from the processing component 22 arranged in the preceding stage, performs an operation on the received image data, and gives it to the processing component 22 arranged in the subsequent stage. For example, the generation unit 50 uses the at least one previous generation image filter including a plurality of filter components as at least one previous generation converter 20 to generate the current generation image filter by genetic processing. In this case, the genetic processing apparatus 10 evolves an image filter group including a plurality of image filters over a plurality of generations based on evolutionary calculation.

  Moreover, the converter 20 may be the structure which combined the processing component 22 which is hardware as an example. In this case, the converter storage part 34 memorize | stores the structure data which show the data transfer relationship between each process components as a data structure showing a converter. The converter 20 may have a configuration in which processing components 22 that are programs for performing operations on data are combined. Further, the converter 20 may have a configuration in which processing components 22 that are arithmetic expressions representing arithmetic contents to be applied to data are combined.

  Further, the converter 20 may convert, for example, a one-dimensional data string, a two-dimensional data group, a three-dimensional data group, or a further multidimensional data group. The one-dimensional data string is, for example, time series data or an array data string. The two-dimensional data group is, for example, image data in which a plurality of pixel data is arranged in a two-dimensional space. The three-dimensional data group is, for example, volume data in which data values representing color or density are arranged at each lattice point in the three-dimensional space. Further, the converter 20 may output data having a dimension different from the input data.

  Each of the plurality of processing components 22 receives input data output from the processing component 22 arranged in the previous stage, performs an operation on the received input data, and gives the processed data to the processing component 22 arranged in the subsequent stage. Each of the plurality of processing components 22 may be a program module, an arithmetic expression, or the like, and performs binarization operation, histogram operation, smoothing operation, edge detection operation, morphology operation, and / or frequency with respect to received input data. Unary operations such as space operations (for example, low-pass filtering operations and high-pass filtering operations) may be performed.

  Each of the plurality of processing components 22 performs an average operation, a difference operation, and / or a fuzzy operation (for example, an OR operation, an AND operation, an algebraic sum, an algebraic product, a limit sum, a limit product, Binomial operations such as intense sum and intense product) may be performed.

  FIG. 4 shows an example of a genetic operation performed on the converter 20 in which the processing components 22 according to this embodiment are combined in series. FIG. 5 shows an example of a crossover operation performed on the converter 20 in which the processing components 22 according to the present embodiment are combined in a tree structure. FIG. 6 shows an example of a mutation operation performed on the converter 20 in which the processing component 22 according to the present embodiment is combined in a tree structure.

  For example, the generation unit 50 performs a crossover operation, which is an example of a genetic operation, on two or more converters 20 to generate two or more new converters 20. As an example, as illustrated in FIGS. 4 and 5, the generation unit 50 converts a part group 24 </ b> A of at least one converter 20 </ b> A that has already been generated to another converter 20 </ b> B that has already been generated. New transducers 20E and 20F are generated by replacing at least a part group 24B. The component group 24 is a member obtained by combining one or a plurality of processing components 22.

  Further, as an example, the generation unit 50 performs a mutation operation, which is an example of a genetic operation, on a single converter 20 to generate a new single converter 20. As an example, as illustrated in FIG. 4 and FIG. 6, the generation unit 50 replaces a part group 24 </ b> C of a part of the already generated one converter 20 </ b> C with another part group 24 </ b> G selected at random. Thus, a new converter 20G is generated.

  The generation unit 50 may leave the current generation converter 20 as the next generation converter 20 as it is. As an example, as illustrated in FIG. 4, the generation unit 50 generates a next-generation converter 20H that includes the configuration of the processing component 22 of the converter 20D as it is.

  The selection unit 46 selects one or a plurality of converters 20 by a technique in which a natural selection of a living organism is modeled with respect to the plurality of converters 20 generated by the generation unit 50. The selection unit 46 may preferentially select the converter 20 having a higher matching degree among the plurality of converters 20. As an example, the selection unit 46 may select the converter 20 according to a technique such as elite selection and roulette selection based on the degree of fitness of each of the plurality of converters 20. Then, the update unit 48 stores the converter 20 in the converter storage unit 34 so that the selected converter 20 can survive to the next generation, and the converter storage unit saves the converter 20 that has not been selected. 34 is deleted.

  FIG. 7 shows an example of an operation flow of the genetic processing apparatus 10 according to the present embodiment. The genetic processing device 10 repeatedly executes each process of step S13 to step S17 a plurality of times (for example, a plurality of generations) (S12, S18). Note that the genetic processing device 10 generates the converter 20 that is randomly generated or has already been generated as the initial converter 20 for generating the converter 20 that converts the given learning input data into the learning target data. Are stored in the converter storage unit 34 in advance.

  In each generation, first, the conversion processing unit 38 generates learning output data obtained by converting the learning input data by the converter 20 for each of the plurality of converters 20 included in the current generation converter group. (S13). The conversion processing unit 38 stores the generated learning output data in the learning output data storage unit 40 in association with the converter 20.

  Subsequently, the fitness level calculation unit 44 calculates the fitness level for each of the plurality of converters 20 included in the current generation transducer group (S14). As an example, the fitness level calculation unit 44 calculates the similarity level or the approximation level between the learning output data and the learning target data as the fitness level. In addition to the degree of approximation, the degree-of-fit calculation unit 44 may calculate the degree of suitability that has been evaluated according to the small number of processing components, the low processing load, the degree of parallelism, and the like. .

  And the collection part 49 collects the information regarding a corresponding generation, for example, when a fitness is more than a reference value (S21). Subsequently, the selection unit 46 preferentially selects the converter 20 having a high fitness among the plurality of converters 20 included in the current generation converter group (S15). For example, the selection unit 46 selects the converter 20 having a higher fitness than the reference value.

  Further, as an example, the selection unit 46 may select a converter 20 in a range in which the fitness is predetermined from the top among the plurality of converters 20 included in the current generation converter group. Moreover, the selection part 46 may select the converter 20 at random on the conditions set so that the converter 20 with higher adaptability may be selected with a higher probability as an example. As an example, the selection unit 46 selects one converter 20 having the highest fitness in the last generation.

  Subsequently, the update unit 48 updates the converter group stored in the converter storage unit 34 (S16). More specifically, the update unit 48 causes the converter 20 selected in step S15 among the plurality of converters 20 included in the current generation converter group to remain in the next generation, and replaces the other converters 20 with each other. The converter group is updated by hesitating. For example, the update unit 48 hesitates by deleting the converter 20 that was not selected in step S15 from the converter storage unit 34.

  Subsequently, the generation unit 50 performs a genetic operation such as crossover and mutation on at least one converter 20 remaining in the converter group by the update process, thereby generating one or more new converters. 20 is generated (S17). In this case, the generation unit 50 performs a genetic operation so that each processing component 22 is introduced to the new converter 20 with the introduction probability set by the collection unit 49. For example, when performing the crossover operation, the generation unit 50 selects one or a plurality of processing components 22 out of a plurality of processing components 22 that can be incorporated in the new converter 20 with a set introduction probability, A part group 24 in the parent converter 20 is replaced with a part group 24 including one or more selected processing parts 22.

  Then, the generation unit 50 stores the remaining converter 20 and the new converter 20 in the converter storage unit 34 as a plurality of converters 20 included in the next-generation converter group. Note that the generation unit 50 does not execute the process in the last generation.

  The weight generation unit 43 generates new weight data for the next generation in order to generate appropriate weight data for each generation (S22). Note that the weight generation unit 43 does not have to execute the process in the last generation. Details of the weight data generation processing will be described later.

  The genetic processing apparatus 10 repeatedly executes the above processing for a plurality of generations (for example, tens of generations or hundreds of generations or more), and performs processing up to the last generation (for example, the Nth generation, where N is a natural number of 2 or more). After the execution, the flow is exited (S18). In this way, the genetic processing device 10 can generate the converter 20 suitable for converting the learning input data into the learning target data using the evolutionary calculation.

  FIG. 8 shows an example of the transition of the fitness level in each generation. The graph of this figure shows the transition of the fitness when the horizontal axis is the generation and the vertical axis is the fitness. The fitness of each generation increases as the converter 20 evolves due to genetic processing by the genetic processing device 10.

  In this example, as an example, the collection unit 49 collects information related to the previous generation when the increase in fitness between the previous generation and the current generation is greater than the reference value. Further, the genetic processing apparatus 10 may use an increment of fitness between two or more generations as an increment of fitness between generations. Here, when the increment of the fitness level is larger than the reference value, it is highly likely that the transducer group including the combination of the processing components 22 that increase the fitness level more greatly exists in the immediately previous generation transducer group.

  Therefore, as an example, when the increase in the fitness between the previous generation and the current generation is larger than the reference value, the collection unit 49 uses the information related to the previous generation converter 20 or the like as the generation information to be collected. Stored in the generation information storage unit 51. The collection unit 49 may store the information related to the current generation converter 20 or the current generation in the generation information storage unit 51 as generation information to be collected.

  As a result, the genetic processing apparatus 10 changes the conversion condition for the generation and makes it possible to re-execute, thereby increasing the possibility that the converter 20 having a higher fitness can be efficiently generated. In addition, the collection unit 49 collects less information when the increase in the fitness between the previous generation and the current generation is smaller than the reference value, compared with the case where the increase in the fitness is larger than the reference value. May be.

  FIG. 9 shows an example of the processing flow in step S21 of FIG. The genetic processing device 10 executes the following processing in step S21 of FIG.

  The collection unit 49 acquires the previous generation and the current generation two generations of fitness calculated by the fitness calculator 44 (S31). Then, the collection unit 49 calculates an increase in fitness between two generations. Here, as an example, the collection unit 49 compares the reference value of the preset increment value with the acquired increment of the fitness between the two generations (S32).

  Subsequently, when the increment of the fitness between the previous generation and the current generation is larger than the reference value (S32: Yes), the collection unit 49, for example, the structure of the elite converter 20 in the previous generation and the conversion At least one of the learning output data obtained by converting the learning input data by the device 20 is collected (S33). When the process of step S33 is completed, or when it is determined in step S32 that the increase in fitness between generations is smaller than the reference value (S32: No), the collection unit 49 exits the process flow. The process proceeds to step S15 in FIG.

  In this way, the genetic processing apparatus 10 can confirm the evolution process by collecting information for each generation. In addition, the genetic processing device 10 enables the user to adjust the generation of the converter 20 having a high fitness by collecting information in detail for a generation that is particularly changing.

  FIG. 10 shows an example in which the collection unit 49 according to the present embodiment records a moving image. For example, the collection unit 49 compares a moving image in which display screens including information about each generation are displayed in the order of generation, with an increase in fitness corresponding to a reference value compared to a case where the increase in fitness between generations is less than or equal to a reference value. Record longer generation time intervals than the value. In addition, the collection unit 49 responds to a case where a moving image that displays a display screen including information on each generation in the order of generation is relatively large compared to other generations in the degree of fitness between generations. Recording may be performed with a longer time interval between generations.

  Here, the collection unit 49 records the generation time interval for which the increment of the fitness is larger than the reference value for a long time, and also records the generation before and after the generation for a longer time than the other generations for which the fitness score is smaller than the reference value. May be. The collection unit 49 may output at least one of an output image output by the elite converter having a fitness value, a weight image in each generation, and an image indicating the structure of the elite converter as an image including information on the generation. Good. In the example of FIG. 10, the collection unit 49 records a generation in which the increment of the fitness is greater than the reference value and the time interval between the two generations before and after the generation.

  For example, the collection unit 49 has 1.0s for generations whose fitness is higher than the reference value, 0.6s for one generation before and after the generation whose fitness is greater than the reference value among generations whose fitness is less than the reference value, Of the generations whose fitness is greater than the reference value are recorded as 0.4 s, and the generations whose other fitness is smaller than the reference value are recorded as 0.2 s.

  As a result, the genetic processing device 10 enables detailed inspection and confirmation by the user of the information on the converter and the change state of the output data for the generation having a large change in the fitness. Then, the genetic processing device 10 can efficiently generate the converter 20 having a higher conversion fitness by the user changing the setting of the introduction probability or the like and re-executing based on the survey result or the like. .

  As described above, according to the genetic processing device 10, for example, the user can easily confirm the state of greater evolution (or change) via the display unit 52. The genetic processing apparatus 10 can redo the evolution by increasing or decreasing the mutation rate and the crossover rate by the user during the course of the evolution. Further, when re-executed, the genetic processing apparatus 10 reliably restores the specified generation of the converter 20, thereby losing the information of the generation before the specified generation, without waste, and with a higher fit. The degree converter 20 can be efficiently generated. Thereby, according to the genetic processing apparatus 10, the target converter 20 can be produced | generated with little calculation cost (including calculation time).

  FIG. 11 shows an example of the processing flow of the weight generation unit 43 in step S22 of FIG. The weight generation unit 43 executes the following processes of steps S131 to S133 in the weight data generation process in step S22 shown in FIG. First, the weight generation unit 43 extracts at least one set of data from the output data and target data output from the converter 20 for generating weight data (S131).

  Subsequently, the weight generation unit 43 calculates a difference for each data between at least one set of data extracted for weight data generation in step S131 (S132). The weight generation unit 43 calculates the difference in data between corresponding positions in the extracted data set. For example, if the extracted data set is a three-dimensional data group, the weight generation unit 43 calculates a data difference between lattice points corresponding to each other. Further, the weight generation unit 43 may calculate the absolute value of the subtraction result of the two data as a difference.

  Further, when the plurality of sets of data are extracted in step S131, for example, the weight generation unit 43 calculates the difference between the sets of data for each data (for example, for each grid point in the case of a three-dimensional data group), Sum or average. Instead of this, the weight generation unit 43 may select, for each data, one largest difference among the differences between the plurality of sets of data.

  Subsequently, the weight generation unit 43 generates the next generation weight data based on the difference for each data calculated in step S132 (S133). More specifically, the weight generation unit 43 generates next-generation weight data that makes the weight of data with a larger difference larger than the weight of data with a smaller difference. The weight generation unit 43 may store weight data generated in advance by the user in the first generation. Further, the weight generation unit 43 may store weight data in which the weight for each data is the same.

  Here, when performing such processing, the weight generation unit 43, in step S131, at least one set of at least one output data and target data output from at least one converter 20 before the current generation. Next-generation weight data is generated from the data.

  For example, the weight generation unit 43 extracts a set of at least one output data selected in descending order of the degree of approximation and target data in at least one or more generations before the current generation. That is, in this case, the weight generation unit 43 extracts a set of output data and target data for each generation before the current generation.

  Further, as an example, the weight generation unit 43 extracts two sets of output data selected in descending order of approximation in at least one generation before the current generation. That is, in this case, the weight generation unit 43 extracts two different sets of output data for each generation before the current generation.

  Further, as an example, the weight generation unit 43 includes a set of at least one output data selected in descending order of approximation and target data in each generation of at least one generation before the current generation, and the degree of approximation. Both sets of two output data selected in descending order are extracted. That is, in this case, the weight generation unit 43 extracts both the set of target data and output data and two different sets of output data for each generation before the current generation.

  The weight generation unit 43 calculates a difference for each data between the data for each set extracted as described above. Then, the weight generation unit 43 generates weight data for the next generation from the difference for each data calculated for each set. Thereby, the weight generation unit 43 weights the difference for each data between each of the two or more output data output from the two or more next-generation converters 20 and the target data by the next-generation weight data. The degree of approximation of each output data can be calculated.

  According to such a genetic processing apparatus 10 according to the present embodiment, the degree of approximation between the output data and the target data can be calculated by appropriately weighting each data with the weight data. Thereby, according to the genetic processing apparatus 10, the output data approximated to target data can be selected appropriately from several output data.

  Furthermore, the genetic processing device 10 generates weight data for the next generation using output data generated in each generation before the current generation. Thus, according to the genetic processing device 10, it is possible to stably generate appropriate weight data in the long term by avoiding a state where the same weight data is periodically generated repeatedly. it can.

  FIG. 12 shows an example of a slide bar for setting introduction probabilities of a plurality of processing components that can be incorporated into the new converter 20. The re-execution unit 53 stores the introduction probability to be introduced into the new converter 20 by genetic operation for each of the processing components 22 that can be incorporated into the converter 20. For example, if the converter 20 is a filter, the re-execution unit 53 stores the introduction probability for each type of filter (for example, for each type such as binarization calculation and average calculation).

  The re-execution unit 53 individually sets and changes the introduction probabilities stored in this way in accordance with external inputs from the user or the like. For example, the re-execution unit 53 displays a slide bar that is moved according to an input from the outside, and each of the plurality of processing components 22 that can be incorporated in at least one converter 20 according to the position of the slide bar. The introduction probability may be changed.

  In the example of FIG. 12, the re-execution unit 53 changes the coefficient corresponding to the ratio of the introduction probability of the corresponding processing component 22 with respect to the average introduction probability of the plurality of processing components 22 with the slide bar. For example, in the example of FIG. 12, when the coefficient is set to 1 by the slide bar, the re-execution unit 53 sets the introduction probability of the corresponding processing component 22 to an average value. For example, the re-execution unit 53 may appropriately change the settings for each of the plurality of processing components 22 from the generation designated by the user on the display unit 52 and re-execute.

  Further, the re-execution unit 53 may set an introduction probability to be introduced into the first generation converter 20 for each of the plurality of processing components 22 that can be incorporated into at least one converter 20. In this case, the generation unit 50 changes the configuration of the converters 20 included in the first generation converter group so that each type of processing component 22 is incorporated at a set ratio.

  As described above, the genetic processing apparatus 10 sets the introduction probability to be introduced into the new converter 20 by genetic operation for each of the plurality of processing components 22. Accordingly, the genetic processing apparatus 10 is configured so that the processing component 22 that is expected to be able to generate the converter 20 having a high fitness by being incorporated is incorporated with a higher probability than the other processing components 22. Can be given a weight. Thereby, the genetic processing apparatus 10 can reduce the processing time and the calculation cost until the target converter 20 is obtained.

  In addition, when the fitness is converged over a plurality of generations, the collection unit 49 uses, as an example, the introduction probability of the processing component 22 estimated to be effective in the application to which the converter 20 is applied as another processing component. It is higher than the introduction probability of 22. As a result, the genetic processing apparatus 10 can re-execute the genetic operation conditions such as mutation during the evolution, so that even if the evolution does not progress, the target converter can be obtained. The possibility that it can be obtained can be increased.

  FIG. 13 shows an example of a hardware configuration of a computer 1900 according to the present embodiment. A computer 1900 according to this embodiment is connected to a CPU peripheral unit having a CPU 2000, a RAM 2020, a graphic controller 2075, and a display device 2080 that are connected to each other by a host controller 2082, and to the host controller 2082 by an input / output controller 2084. An input / output unit having a communication interface 2030, a hard disk drive 2040, and a DVD drive 2060, and a legacy input / output unit having a ROM 2010, a flexible disk drive 2050, and an input / output chip 2070 connected to the input / output controller 2084 Is provided.

  The host controller 2082 connects the RAM 2020 to the CPU 2000 and the graphic controller 2075 that access the RAM 2020 at a high transfer rate. The CPU 2000 operates based on programs stored in the ROM 2010 and the RAM 2020 and controls each unit. The graphic controller 2075 acquires image data generated by the CPU 2000 or the like on a frame buffer provided in the RAM 2020 and displays it on the display device 2080. Instead of this, the graphic controller 2075 may include a frame buffer for storing image data generated by the CPU 2000 or the like.

  The input / output controller 2084 connects the host controller 2082 to the communication interface 2030, the hard disk drive 2040, and the DVD drive 2060, which are relatively high-speed input / output devices. The communication interface 2030 communicates with other devices via a network. The hard disk drive 2040 stores programs and data used by the CPU 2000 in the computer 1900. The DVD drive 2060 reads a program or data from the DVD 2095 and provides it to the hard disk drive 2040 via the RAM 2020.

  The input / output controller 2084 is connected to the ROM 2010, the flexible disk drive 2050, and the relatively low-speed input / output device of the input / output chip 2070. The ROM 2010 stores a boot program that the computer 1900 executes at startup and / or a program that depends on the hardware of the computer 1900. The flexible disk drive 2050 reads a program or data from the flexible disk 2090 and provides it to the hard disk drive 2040 via the RAM 2020. The input / output chip 2070 connects the flexible disk drive 2050 to the input / output controller 2084 and inputs / outputs various input / output devices via, for example, a parallel port, a serial port, a keyboard port, a mouse port, and the like. Connect to controller 2084.

  A program provided to the hard disk drive 2040 via the RAM 2020 is stored in a recording medium such as the flexible disk 2090, the DVD 2095, or an IC card and provided by the user. The program is read from the recording medium, installed in the hard disk drive 2040 in the computer 1900 via the RAM 2020, and executed by the CPU 2000.

  A program installed in the computer 1900 and causing the computer 1900 to function as the genetic processing apparatus 10 includes a conversion processing module, a weight generation module, a fitness calculation module, a selection module, an update module, a collection module, and a generation module. And a re-execution module. These programs or modules work on the CPU 2000 or the like to make the computer 1900, the conversion processing unit 38, the weight generation unit 43, the fitness calculation unit 44, the selection unit 46, the update unit 48, and the collection unit 49. , Function as a generation unit 50 and a re-execution unit 53, respectively.

  The information processing described in these programs is read into the computer 1900, whereby the conversion processing unit 38, which is a specific means in which the software and the various hardware resources described above cooperate, the weight generation unit 43, , The degree-of-fit calculation unit 44, the selection unit 46, the update unit 48, the collection unit 49, the generation unit 50, and the re-execution unit 53. And the specific genetic processing apparatus 10 according to the intended purpose is constructed | assembled by implement | achieving the calculation or processing of the information according to the intended purpose of the computer 1900 in this embodiment by these specific means.

  As an example, when communication is performed between the computer 1900 and an external device or the like, the CPU 2000 executes a communication program loaded on the RAM 2020 and executes a communication interface based on the processing content described in the communication program. A communication process is instructed to 2030. Under the control of the CPU 2000, the communication interface 2030 reads transmission data stored in a transmission buffer area or the like provided on a storage device such as the RAM 2020, the hard disk drive 2040, the flexible disk 2090, or the DVD 2095, and transmits it to the network. Alternatively, the reception data received from the network is written into a reception buffer area or the like provided on the storage device. As described above, the communication interface 2030 may transfer transmission / reception data to / from the storage device by a DMA (direct memory access) method. Instead, the CPU 2000 transfers the storage device or the communication interface 2030 as a transfer source. The transmission / reception data may be transferred by reading the data from the data and writing the data to the transfer destination 2030 or the storage device.

  In addition, the CPU 2000 reads all or necessary portions from files or databases stored in an external storage device such as the hard disk drive 2040, the DVD drive 2060 (DVD 2095), and the flexible disk drive 2050 (flexible disk 2090). The data is read into the RAM 2020 by DMA transfer or the like, and various processes are performed on the data on the RAM 2020. Then, CPU 2000 writes the processed data back to the external storage device by DMA transfer or the like. In such processing, since the RAM 2020 can be regarded as temporarily holding the contents of the external storage device, in the present embodiment, the RAM 2020 and the external storage device are collectively referred to as a memory, a storage unit, or a storage device. Various types of information such as various programs, data, tables, and databases in the present embodiment are stored on such a storage device and are subjected to information processing. Note that the CPU 2000 can also store a part of the RAM 2020 in the cache memory and perform reading and writing on the cache memory. Even in such a form, the cache memory bears a part of the function of the RAM 2020. Therefore, in the present embodiment, the cache memory is also included in the RAM 2020, the memory, and / or the storage device unless otherwise indicated. To do.

  In addition, the CPU 2000 performs various operations, such as various operations, information processing, condition determination, information search / replacement, etc., described in the present embodiment, specified for the data read from the RAM 2020 by the instruction sequence of the program. Is written back to the RAM 2020. For example, when performing the condition determination, the CPU 2000 determines whether the various variables shown in the present embodiment satisfy the conditions such as large, small, above, below, equal, etc., compared to other variables or constants. When the condition is satisfied (or not satisfied), the program branches to a different instruction sequence or calls a subroutine.

  Further, the CPU 2000 can search for information stored in a file or database in the storage device. For example, in the case where a plurality of entries in which the attribute value of the second attribute is associated with the attribute value of the first attribute are stored in the storage device, the CPU 2000 displays the plurality of entries stored in the storage device. The entry that matches the condition in which the attribute value of the first attribute is specified is retrieved, and the attribute value of the second attribute that is stored in the entry is read, thereby associating with the first attribute that satisfies the predetermined condition The attribute value of the specified second attribute can be obtained.

  The program or module shown above may be stored in an external recording medium. As the recording medium, in addition to the flexible disk 2090 and the DVD 2095, an optical recording medium such as DVD or CD, a magneto-optical recording medium such as MO, a tape medium, a semiconductor memory such as an IC card, and the like can be used. Further, a storage device such as a hard disk or RAM provided in a server system connected to a dedicated communication network or the Internet may be used as a recording medium, and the program may be provided to the computer 1900 via the network.

  As mentioned above, although this invention was demonstrated using embodiment, the technical scope of this invention is not limited to the range as described in the said embodiment. It will be apparent to those skilled in the art that various modifications or improvements can be added to the above-described embodiment. It is apparent from the scope of the claims that the embodiments added with such changes or improvements can be included in the technical scope of the present invention.

  The order of execution of each process such as operations, procedures, steps, and stages in the apparatus, system, program, and method shown in the claims, the description, and the drawings is particularly “before” or “prior to”. It should be noted that the output can be realized in any order unless the output of the previous process is used in the subsequent process. Regarding the operation flow in the claims, the description, and the drawings, even if it is described using “first”, “next”, etc. for convenience, it means that it is essential to carry out in this order. It is not a thing.

DESCRIPTION OF SYMBOLS 10 Genetic processor, 20 Converter, 22 Processing components, 24 Parts group, 34 Converter storage part, 36 Learning input data storage part, 38 Conversion processing part, 40 Learning output data storage part, 42 Learning target data Storage unit, 43 Weight generation unit, 44 Fitness calculation unit, 46 Selection unit, 48 Update unit, 49 Collection unit, 50 generation unit, 51 Generation information storage unit, 52 Display unit, 53 Re-execution unit, 1900 Computer, 2000 CPU 2010 ROM, 2020 RAM, 2030 communication interface, 2040 hard disk drive, 2050 flexible disk drive, 2060 DVD drive, 2070 input / output chip, 2075 graphic controller, 2080 display device, 2082 host controller, 2084 input / output controller Controller, 2090 flexible disk, 2095 DVD

Claims (13)

A generation unit for generating a current generation converter by genetic processing from at least one previous generation converter including a plurality of processing components that process input data and output a processing result as output data;
For each of the current generation converters, a fitness calculation unit that calculates the fitness for conversion from learning input data to learning target data; and
A collecting unit that collects information about a corresponding generation when the increment of the fitness between generations is larger than a reference value or when the increment of the fitness between generations is relatively larger than other generations. When,
With
The collection unit displays a moving image that displays a display screen including information on the corresponding generation in each generation in the order of generation, and the time interval of the generation in which the increment of the fitness is greater than the reference value or the increment of the fitness is A genetic processing device that records a longer generation time interval than other generations. The genetic processing device according to claim 1, wherein the collection unit collects more information regarding the corresponding generation as compared with a case where the increment of the fitness between generations is equal to or less than the reference value. It said collecting unit, as the information on the corresponding generation, according to claim 1 or 2 for collecting at least one of said transducer structures and converted for learning output data input data for the learning by the transducer Genetic processing equipment. The collection unit collects at least one of the structure of the converter and the output data for learning obtained by converting the learning input data by the converter as information on the corresponding generation according to the fitness. Item 4. The genetic processing apparatus according to Item 3 . Said collecting unit, said as the information for the corresponding generation, according to claim 3 or 4 for collecting at least one of the superior transducer structure and the excellent output data for learning obtained by converting the input data for the learning by the transducer was The genetic processing device according to 1. A weight generation unit that updates the weight data representing the weight of each data area for each generation,
The fitness calculation unit calculates, for each of the current generation converters, the fitness for conversion from the learning input data to the learning target data by weighting the weight data,
When the increment of the fitness level between generations is larger than the reference value or when the increment of the fitness level between generations is relatively large compared to other generations, the collection unit The genetic processing device according to any one of claims 1 to 5 , wherein at least the weight data is collected as information regarding. The genetic processing device according to claim 1, further comprising: a display unit configured to display information on the corresponding generation in the designated one generation when a generation is designated during the output of the moving image. When the increment of the fitness between generations is larger than the reference value or when the increment of the fitness between generations is relatively large compared to other generations, all the converters The genetic processing device according to any one of claims 1 to 7 , wherein the genetic processing device is stored reproducibly. Re-execution to reproduce all the converters according to designation from the outside, set processing conditions different from the processing conditions for generating all the converters for the generation unit, and re-execute genetic processing The genetic processing device according to claim 8 . The re-execution unit, as the processing conditions, the mutation rate by the generating unit, the crossover rate, and, in claim 9, each of the plurality of processing parts are set at least one of the introduction probability of being introduced to the transducer The genetic processor described. The generation unit uses a plurality of filter components that respectively convert an input image to an output image as the plurality of processing components, and converts at least one previous generation image filter including the plurality of filter components into the at least one previous generation image filter. The genetic processing device according to any one of claims 1 to 10, wherein a current generation image filter is generated as a converter by genetic processing. Generating a current generation converter by genetic processing from at least one previous generation converter including a plurality of processing components that process input data and output processing results as output data;
For each of the current generation converters, a fitness calculation step for calculating a fitness for conversion from learning input data to learning target data; and
A collection step of collecting information about a corresponding generation when the increase in fitness between generations is greater than a reference value or when the increase in fitness between generations is relatively large compared to other generations. When,
With
In the collecting step, a moving image displaying a display screen including information on the corresponding generation in each generation in the order of generation, a time interval of generations in which the fitness increment is larger than the reference value or the fitness increment is calculated. A genetic processing method in which a relatively large generation time interval is set longer than that of other generations. A genetic processing program for causing a computer to function as a genetic processing device,
The computer,
A generation unit for generating a current generation converter by genetic processing from at least one previous generation converter including a plurality of processing components that process input data and output a processing result as output data;
For each of the current generation converters, a fitness calculation unit that calculates the fitness for conversion from learning input data to learning target data; and
A collecting unit that collects information about a corresponding generation when the increment of the fitness between generations is larger than a reference value or when the increment of the fitness between generations is relatively larger than other generations. When,
With
The collection unit displays a moving image in which display screens including information on the corresponding generations in each generation are displayed in generation order, a time interval of generations where the fitness increment is greater than the reference value, or the fitness increment is A genetic processing program that functions as a genetic processing device that records a longer generation time interval than other generations.

Images