Intelligent meeting assistance system and method of generating meeting minutes

Abstract

To provide, in order to resolve the inadequacies of existing techniques, an intelligent meeting assistance system capable of analyzing images containing text and / or charts to generate meeting minutes that record content of the images, and to provide a method of generating meeting minutes.SOLUTION: The present invention relates to an intelligent meeting assistance system and a method of generating meeting minutes. The intelligent meeting assistance system includes an image capturing device and an image analysis device. The image capturing device is configured to capture an image displayed by an interactive device during a meeting. The image analysis device is coupled to the image capturing device and is configured to execute an image analysis process on the image to generate first meeting minutes that record image content.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to an auxiliary system, and more particularly to an intelligent conference support system and a method for generating a conference record.

Background Art

[0002] Basically, an intelligent conference support system is a conference assistant that automatically generates a conference record. However, existing intelligent conference support systems only generate a conference record that records multiple voice contents based on voice data. However, humans are visual learners.

[0003] Therefore, a conference presenter can use an image including text and / or a chart to explain information to conference participants. That is, existing intelligent conference support systems ignore image content other than voice, and it will not be reflected in the conference record.

Summary of the Invention

Problems to be Solved by the Invention

[0004] The technical problem to be solved by the present invention is to provide an intelligent conference support system and a method for generating a conference record that analyze an image including text and / or a chart and generate a conference record that records the image content in order to make up for the deficiencies of the existing technology.

Means for Solving the Problems

[0005] In order to solve the above technical problem, one of the technical solutions adopted by the present invention is to provide an intelligent conference support system including an image capture device and an image analysis device. The image capture device is configured to capture an image displayed by an interaction device during a conference. The image analysis device is coupled to the image capture device and configured to perform an image analysis process on the image and generate a first conference record that records the image content.

[0006] To solve the above technical problem, another technical solution adopted by the present invention is to set an image capture device for the image capture device to capture the image displayed by the interaction device during the meeting period, and the image analysis device to set the image analysis device to execute an image analysis process on the image and generate a first meeting record recording the image content. A method for generating a meeting record is provided.

[0007] To better understand the features and technical content of the invention, the following refers to the detailed description of the present invention and the accompanying drawings. However, the provided accompanying drawings are only for reference and explanation, and are not for limiting the scope of the claims of the present invention.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3

Figure 4A

Figure 4B

Figure 4C

Figure 5

Figure 6A

Figure 6B

Figure 6C

Embodiments for Carrying Out the Invention

[0009] The embodiments disclosed by the present invention will be described below. Those skilled in the art can understand the merits and effects of the present invention from the disclosure of this specification. The present invention can be implemented or applied by other different embodiments. Each detail in this specification can also be equally modified and changed based on various viewpoints or applications without departing from the spirit of the present invention. Also, the drawings of the present invention are for simple and schematic explanation and do not show actual dimensions. In the following embodiments, the technical matters related to the present invention will be further described, but the disclosed content does not limit the present invention. Also, the term "or" used in this specification can include any one or a combination of multiple items according to the actual situation.

[0010] Please refer to FIGS. 1 and 2 together. FIG. 1 is a functional block diagram of the intelligent conference support system according to an embodiment of the present invention, and FIG. 2 is a flowchart showing the steps of a method for generating a conference record according to an embodiment of the present invention. As shown in FIG. 1, the intelligent conference support system 1 of the present embodiment includes an image capture device 11 and an image analysis device 12. Specifically, the image capture device 11 is configured to capture the image 4 displayed by the interaction device 2 during the conference. Also, the image analysis device 12 is communicatively connected to the image capture device 11, performs an image analysis process on the image 4, and is configured to generate a first conference record M1 for recording the image content (not shown in FIG. 1).

[0011] For illustration purposes, the interactive device 2 is a touch screen within the conference environment 3, configured to display slides including text and / or charts during the conference, enabling the conference presenter 5 to explain information to conference participants (not shown in FIG. 1). In this case, the image 4 captured by the image capture device 11 is a slide including text and / or charts, and the image analysis device 12 can perform an image analysis process on the slide including text and / or charts to generate a first meeting record M1 that records the image content of the slide. However, the present invention is not limited to the above example.

[0012] As shown in FIG. 2, based on the above content, the method for generating a meeting record according to this embodiment is executed by the intelligent conference support system 1 and includes the following steps.

[0013] Step S11: Set the image capture device to capture the image displayed by the interactive device during the conference. Specifically, the image capture device 11 can be realized by combining hardware (e.g., lenses and imaging media) with software and / or firmware. However, the present invention is not limited to the specific implementation form of the image capture device 11.

[0014] Step S12: Set the image analysis device to perform an image analysis process on the image and generate a first meeting record that records the image content. Similarly, the image analysis device 12 can be realized by combining hardware (e.g., a central processing unit and memory) with software and / or firmware. However, the present invention is not limited to the specific implementation form of the image analysis device 12.

[0015] Furthermore, the image content recorded in the first meeting record M1 includes either the text content or the chart content in Image 4, or a combination thereof, and the image analysis process performed by the image analysis apparatus 12 includes a text recognition process and a chart recognition process. Therefore, as shown in FIG. 1, the image analysis apparatus 12 includes an optical character recognition circuit 121 and a chart recognition circuit 122.

[0016] The optical character recognition circuit 121 is arranged to perform a text recognition process on Image 4 and generate a first event record for recording the text content. However, the first event record is not shown in FIG. 1. Also, the chart recognition circuit 122 is arranged to perform a chart recognition process on Image 4 and generate a second event record for recording the chart content. Similarly, the second event record is not shown in FIG. 1. That is, please refer to FIG. 3. FIG. 3 is a flowchart of the steps for the image analysis apparatus to perform an image analysis process on an image according to an embodiment of the present invention.

[0017] As shown in FIG. 3, based on the above description, step S12 of this embodiment can include the following steps.

[0018] Step S121: Set the optical character recognition circuit to perform a text recognition process on the image and generate a first event record for recording the text content. Specifically, the text in Image 4 can include either printed text typeset in a modern computer font and handwritten text written by the conference presenter 5 on the interactive device 2 (e.g., touch screen), or a combination thereof. Therefore, the text content recorded in the first event record includes either the printed text content or the handwritten text content in Image 4, or a combination thereof, and the optical character recognition circuit 121 performs a text recognition process to recognize the printed text content and the handwritten text content on Image 4 and generate a first event record for recording the text content.

[0019] Step S122: Set the chart recognition circuit to perform a chart recognition process on the image and generate a second event record that records the chart content. Specifically, the chart in Image 4 can include any one or a combination of a pie chart, a line chart, and a bar chart. Therefore, the chart content recorded in the second event record includes any one or a combination of the pie chart content, the line chart content, and the bar chart content in Image 4. The chart recognition circuit 122 recognizes the pie chart content, the line chart content, and the bar chart content in Image 4 and performs a chart recognition process to generate a second event record that records the chart content.

[0020] Similarly, the pie chart, line chart, and / or bar chart in Image 4 are not only created using a computer's charting application but can also be drawn by the conference presenter 5 on the interactive device 2 (e.g., a touch screen). That is, based on the above content, the intelligent conference support system 1 and the method for generating a conference record in this embodiment can capture and record the content written or drawn by the conference presenter 5 on the interactive device 2. Therefore, the generated conference record can reflect all the actions of the conference in more detail.

[0021] On the other hand, the conference presenter 5 can also interact with conference participants using the motion display function on Image 4 (e.g., a slide). The motion display function is a function that performs operations such as moving, rotating, appearing / disappearing on text content and chart content using various trajectories and / or various screen switching effects. Therefore, the image analysis process performed by the image analysis device 12 also includes a motion display recognition process, and the image content recorded in the first conference record M1 can include any one or a combination of the text content, the chart content, and the motion display content on Image 4. Thus, as shown in FIG. 1, the image analysis device 12 can also include a motion display recognition circuit 123.

[0022] The motion display recognition circuit 123 is arranged to perform a motion display recognition process on the image 4 and generate a third event record for recording the motion display content. In certain embodiments, since the related content of the above-described motion display is usually the key points of the meeting that the meeting presenter 5 wants to emphasize, in the meeting record generation method provided by the present invention, the recognized motion display content is displayed in the third event record as the key part of the meeting (for example, by a weighting method). However, the third event record is not shown in FIG. 1. Also, the image analysis device 12 is also set to generate a first meeting record M1 for recording the image content based on the first event record, the second event record, and the third event record. That is, as shown in FIG. 3, step S12 of this embodiment may include the following steps.

[0023] Step S123: Arrange the motion display recognition circuit to execute a motion display recognition process on the image and generate a third event record for recording the motion display content.

[0024] Step S124: Set the image analysis device to generate a first meeting record for recording the image content based on the first event record, the second event record, and / or the third event record.

[0025] It should be noted that the present invention is not limited to the specific form of the image content recorded in the first meeting record M1. For example, the first meeting record M1 of this embodiment can record the text content, chart content, and motion display content on the image 4 in text form. Also, the first meeting record M1 of this embodiment can also record the chart content on the image 4 in the form of recreation. However, the present invention is not limited to the above examples.

[0026] On the one hand, as shown in FIG. 1, the intelligent conference support system 1 of this embodiment can include a voice input device 13 and a voice analysis device 14. The voice input device 13 is arranged to generate voice data during the conference, but the voice data is not shown in FIG. 1. Further, the voice analysis device 14 is coupled to the voice input device 13 and is arranged to execute a voice analysis process on the voice data, generating a second conference record M2 that records a plurality of voice contents. That is, as shown in FIG. 2, the method of this embodiment can include the following steps.

[0027] Step S13: Set a voice input device to generate voice data during the conference. Specifically, the voice input device 13 is arranged to execute a recording process during the conference to generate voice data. Also, the voice input device 13 may be realized by a combination of hardware (e.g., a microphone) and software and / or firmware. However, the present invention is not limited to the specific realization method of the voice input device 13.

[0028] Step S14: Set a voice analysis device to execute a voice analysis process on the voice data to generate a second conference record that records a plurality of voice contents. Specifically, the voice analysis process executed by the voice analysis device 14 can include a voice recognition process and a voice signature identification process. Accordingly, the voice analysis device 14 can include a voice recognition circuit 141 and a voice signature identification circuit 142, which execute the voice recognition process and the voice signature identification process respectively, generating a second conference record M2 that records a plurality of voice contents.

[0029] Similarly, the voice recognition circuit 141 and the voice signature identification circuit 142 may be realized by a combination of hardware (e.g., a central processing unit and a memory) and software and / or firmware. However, the present invention is not limited to the specific realization method of the voice recognition circuit 141 and the voice signature identification circuit 142. Since the operating principles of the voice recognition circuit 141 and the voice signature identification circuit 142 are known to those skilled in the art, the details of the voice analysis device 14 will not be described further herein.

[0030] Furthermore, the conference presenter 5 may explain information to the conference participants using a plurality of images 4 (for example, a plurality of slides) during the conference. Therefore, the image capture device 11 of the present embodiment is arranged to sequentially capture the plurality of images 4 displayed on the interaction device 2 during the conference, and the image analysis device 12 generates a first conference record M1 that records each of the plurality of image contents. It should be understood that the plurality of image contents recorded in the first conference record M1 respectively correspond to the plurality of images 4 captured by the image capture device 11.

[0031] In the present embodiment, the image capture device 11 is configured to sequentially capture a plurality of images 4 based on a sampling frequency, but the present invention is not limited thereto. In other embodiments, the image capture device 11 is configured to capture a new image 4 when the image 4 is updated (for example, when the conference presenter 5 switches to another slide to explain information to the conference participants, or when the conference presenter 5 writes text or draws a chart on the interaction device 2), but the present invention is not limited to this either.

[0032] Also, each image content should be understood to be recorded in a second conference record M2 associated with at least one voice content. Therefore, in order to establish the relevance and order of the plurality of image contents and the plurality of voice contents, the image analysis device 12 is configured to add a time stamp to each image content recorded in the first conference record M1, and the voice analysis device 14 is configured to add a time stamp to each voice content recorded in the second conference record M2.

[0033] Based on the above content, since each image content recorded in the first meeting record M1 can include any one of text content, chart content, and motion display content, or a combination thereof, adding a timestamp to each image content means adding a timestamp to the text content, chart content, and / or motion display content of each image content. Therefore, the intelligent conference support system 1 of this embodiment can also include a clock circuit 15.

[0034] The clock circuit 15 is coupled to the image analysis device 12 and the voice analysis device 14 and is used for generating timestamps. That is, the first meeting record M1 and the second meeting record M2 generated by the image analysis device 12 and the voice analysis device 14 can use the timestamps generated by the clock circuit 15 to record the occurrence times of each image content and each voice content.

[0035] Furthermore, the intelligent conference support system 1 of this embodiment can also include an artificial intelligence processing circuit 16. The artificial intelligence processing circuit 16 is coupled to the image analysis device 12 and the voice analysis device 14 and receives the first meeting record M1 and the second meeting record M2. Specifically, the artificial intelligence processing circuit 16 integrates the first meeting record M1 and the second meeting record M2, and is configured to input the integrated meeting record into a natural language processing (NLP) and machine learning model to generate an analyzed and comprehensive third meeting record M3.

[0036] In other words, the intelligent conference support system 1 of this embodiment has the ability to generate the first meeting record M1 and the second meeting record M2 that record image content and voice content separately by combining text recognition, chart recognition, and voice recognition, and integrate the first meeting record M1 and the second meeting record M2 through timestamps. And the intelligent conference support system 1 of this embodiment can further analyze the integrated meeting record using natural language processing and machine learning models.

[0037] Basically, through natural language processing and machine learning models, the artificial intelligence processing circuit 16 can better understand the image content and audio content recorded in the first meeting record M1 and the second meeting record M2, correct and supplement the content of recording errors and incomplete records, and then extract and organize the context information, thereby generating a comprehensive third meeting record M3. That is, the third meeting record M3 records all the image content and audio content, and all the image content and audio content are reflected in the third meeting record M3 in order. Based on the above content, as shown in FIG. 2, the method of this embodiment can also include the following steps.

[0038] Step S15: Arrange the artificial intelligence processing circuit to integrate the first meeting record and the second meeting record, input the integrated meeting record into the natural language processing and machine learning models, and analyze and generate the third meeting record.

[0039] Hereinafter, the method for the artificial intelligence processing circuit 16 to correct and supplement the content of recording errors and incomplete records will be further described. However, the present invention is not limited to the following examples. For example, a certain image content recorded in the first meeting record M1 may be the handwritten text content of "U8B", but in fact, it is the handwritten text content that the meeting presenter 5 wrote "USB" on the interactive device 2. That is, there is an error in text recognition, and incorrect text content is recorded in the first meeting record M1.

[0040] Next, based on the time stamp, when the artificial intelligence processing circuit 16 integrates the first meeting record M1 and the second meeting record M2, it can find at least one audio content associated with the aforementioned image content. This at least one audio content is that the meeting presenter 5 is explaining the Universal Serial Bus to the meeting participants. Therefore, through the natural language processing and machine learning models, the artificial intelligence processing circuit 16 can understand that the aforementioned handwritten text content should be "USB" rather than "U8B", and can further correct the aforementioned handwritten text content.

[0041] On the one hand, the other image content recorded in the first meeting record M1 may be the content of the pie chart of the "support rate of candidates". That is, at this point, there is a pie chart reflecting the "support rate of candidates" on Image 4, but the chart recognition circuit 122 may not be able to accurately recognize which candidate's support rate each sector on the corresponding pie chart represents. As a result, incomplete pie chart content will be recorded in the first meeting record M1.

[0042] Similarly, based on the timestamp, the artificial intelligence processing circuit 16 can find at least one audio content associated with the aforementioned image content. This at least one audio content is what the meeting presenter 5 explains to the meeting participants which candidate's support rate each sector on the corresponding pie chart represents. Therefore, through natural language processing and machine learning models, the artificial intelligence processing circuit 16 can supplement the incomplete aforementioned pie chart content.

[0043] Furthermore, the artificial intelligence processing circuit 16 is configured to output a real-time third meeting record M3. Therefore, the intelligent meeting support system 1 of this embodiment also includes an output device 17. The output device 17 is coupled to the artificial intelligence processing circuit 16 and is configured to store and / or display the third meeting record M3. For example, the output device 17 can be a smartphone, a notebook computer, an external storage device, or a set-top box, but the present invention is not limited thereto. Similarly, the present invention is not limited to the specific forms of the image content and audio content recorded in the third meeting record M3.

[0044] Next, the implementation of the optical character recognition circuit 121, the chart recognition circuit 122, and the motion display recognition circuit 123 will be described through specific embodiments below, but the present invention is not limited thereto. Please refer to FIGS. 4A to 4C. FIG. 4A is a functional block diagram of the optical character recognition circuit according to an embodiment of the present invention, FIG. 4B is a functional block diagram of the chart recognition circuit according to an embodiment of the present invention, and FIG. 4C is a functional block diagram of the motion display recognition circuit according to an embodiment of the present invention.

[0045] As shown in FIG. 4A, the optical character recognition circuit 121 can include a Picture Selection circuit 1211, a Picture Segmentation circuit 1212, a Picture Reconstruct circuit 1213, and an optical character recognition engine 1214, and is used to generate a first event record of the recorded text content. However, the first event record is not shown in FIG. 4A. Since the application principle of text recognition is already known to those skilled in the art, the details of the Picture Selection circuit 1211, the Picture Segmentation circuit 1212, the Picture Reconstruct circuit 1213, and the optical character recognition engine 1214 will not be described in further detail.

[0046] It should be noted that since the image analysis device 12 can add time stamps to the text content, chart content, and / or motion display content of each image content, the optical character recognition circuit 121 can also include a time stamp - event synthesis circuit 1215. The time stamp - event synthesis circuit 1215 is coupled to the optical character recognition engine 1214 and is arranged to add a time stamp to the text content recorded in the first event record.

[0047] As shown in FIG. 4B, the chart recognition circuit 122 can include an image selection circuit 1221, an image segmentation circuit 1222, and a chart recognition engine 1223, and is used to generate a second event record of the recorded chart content. However, the second event record is not shown in FIG. 4B. Since the application principle of chart recognition is already known to those skilled in the art, the details of the image selection circuit 1221, the image segmentation circuit 1222, and the chart recognition engine 1223 will not be described in further detail.

[0048] It should be noted that since the chart recognition circuit 122 recognizes the circular chart content, line chart content, and bar chart content on Image 4, the chart recognition engine 1223 can further include a circular chart recognition engine 12231, a line chart recognition engine 12232, and a bar chart recognition engine 12233. Since the application principles of circular chart recognition, line chart recognition, and bar chart recognition are already known to those skilled in the art, the details of the circular chart recognition engine 12231, the line chart recognition engine 12232, and the bar chart recognition engine 12233 will not be described in further detail.

[0049] Similarly, since the image analysis device 12 can add time stamps to the text content, chart content, and / or motion display content of each image content, the chart recognition circuit 122 can also include a time stamp - event synthesis circuit 1224. The time stamp - event synthesis circuit 1224 is coupled to the chart recognition engine 1223 and is arranged to add time stamps to the chart content recorded in the second event record.

[0050] As shown in FIG. 4C, the motion display recognition circuit 123 can include an image segmentation circuit 1231 and a motion display recognition engine 1232, and is used to generate a third event record of the recorded motion display content. However, the third event record is not shown in FIG. 4C. Since the application principle of motion display recognition is already known to those skilled in the art, the details of the image segmentation circuit 1231 and the motion display recognition engine 1232 will not be described in further detail.

[0051] Similarly, since the image analysis device 12 can add time stamps to the text content, chart content, and / or motion display content of each image content, the motion display recognition circuit 123 can also include a time stamp-event synthesis circuit 1233. The time stamp-event synthesis circuit 1233 is coupled to the motion display recognition engine 1232 and is arranged to add a time stamp to the motion display content recorded in the third event record. It should be noted again that the present invention is not limited to the specific embodiments of the optical character recognition circuit 121, the chart recognition circuit 122, and the motion display recognition circuit 123.

[0052] On the other hand, the following will be described through specific embodiments for explaining the embodiments of the voice recognition circuit 141, but the present invention is not limited thereto. Please refer to FIG. 5. FIG. 5 is a functional block diagram of the voice recognition circuit according to an embodiment of the present invention.

[0053] As shown in FIG. 5, the voice recognition circuit 141 can include a voice preprocessing circuit 1411, a voice recognition engine 1412, a cloud recognition unit 1413, and a local recognition unit 1414. Since the application principle of voice recognition is already known to those skilled in the art, the details of the voice preprocessing circuit 1411, the voice recognition engine 1412, the cloud recognition unit 1413, and the local recognition unit 1414 will not be described in further detail.

[0054] Similarly, the speech recognition circuit 141 can also include a timestamp event synthesis circuit 1415. The timestamp event synthesis circuit 1415 is coupled to the speech recognition engine 1412 and is arranged to add timestamps to the speech content.

[0055] Furthermore, the intelligent conference support system 1 can display speech content, text content, and chart content in a copy mode or an overview mode. Refer to FIG. 6A. FIG. 6A is a schematic diagram showing that the intelligent conference support system according to an embodiment of the present invention displays speech content in the copy mode. As shown in FIG. 6A, since each speech content can include a text, in the copy mode, a plurality of texts (for example, from Sent1 to Sent5) are displayed on the screen of the output device 17 in real time and arranged in chronological order.

[0056] Furthermore, when there are a plurality of speakers, the speech recognition circuit 141 can also identify the speaker of each text. For example, each text in the present embodiment corresponds to a speaker Sp1 or Sp2, and the speakers Sp1 and Sp2 can be a conference presenter 5 and a certain conference participant, but the present invention is not limited thereto. Therefore, as shown in FIG. 6A, in the copy mode, the intelligent conference support system 1 can also display the corresponding speaker for each text.

[0057] Next, refer to FIG. 6B. FIG. 6B is a schematic diagram showing that the intelligent conference support system according to an embodiment of the present invention displays speech content in the overview mode. As shown in FIG. 6B, in the overview mode, the intelligent conference support system 1 can display all texts by the same speaker. Also, as described above, the speech analysis device 14 is configured to add timestamps to each speech content. Therefore, the texts Sent1 to Sent5 in the present embodiment respectively correspond to timestamps St1 to St5, and in the overview mode, the intelligent conference support system 1 can also display the timestamp of each text.

[0058] Furthermore, the artificial intelligence processing circuit 16 can generate a more concise quick summary by editing, aggregating, customizing, and optimizing the meeting minutes. Therefore, in the overview mode, the intelligent meeting support system 1 can also display the quick summary QS generated by the artificial intelligence processing circuit 16, thereby helping the user review the entire meeting process.

[0059] On the other hand, compared with the quick summary, the artificial intelligence processing circuit 16 can also generate a more complete summary of the meeting. Therefore, the intelligent meeting support system 1 can display the summary of the meeting generated by the artificial intelligence processing circuit 16 in the summary mode. Please refer to FIG. 6C. FIG. 6C is a schematic diagram showing the intelligent meeting support system according to an embodiment of the present invention displaying the summary of the meeting in the summary mode.

[0060] As shown in FIG. 6C, the artificial intelligence processing circuit 16 can generate a summary of the meeting including a title MS1, summary content MS2, keywords MS3, pain points MS4, action items MS5, and a chart MS6 by editing, aggregating, and optimizing the meeting minutes. Therefore, in the summary mode, the title MS1, summary content MS2, keywords MS3, pain points MS4, action items MS5, and chart MS6 of the meeting summary can be displayed on the screen of the output device 17. In this embodiment, the chart MS6 of the meeting summary is an example of a pie chart, but the present invention is not limited thereto.

[0061] Furthermore, the output device 17 can provide a graphical user interface including a check box B1, and the user can determine whether to display the keywords MS3, pain points MS4, operation items MS5, and charts MS6 of the meeting summary. In addition, the graphical user interface provided by the output device 17 can also include a button B2 with the text "Chart Conversion". When the button B2 is pressed, the user can convert the type of the chart MS6. Similarly, the graphical user interface provided by the output device 17 can also include buttons B3, B4, and B5 with the text "Edit", "Email", and "Save". When the button B3, B4, or B5 is pressed, the user can edit the meeting summary, send the meeting summary by email, or save the meeting summary.

[0062] As described above, one beneficial effect of the present invention is to provide an intelligent meeting support system and a method for generating meeting records, thereby having the ability to generate meeting records of image content through technical means of "capturing images displayed by the interaction device during the meeting" and "performing an image analysis process on the images".

[0063] Furthermore, the intelligent meeting support system and the method for generating meeting records provided by the present invention have the ability to record image content other than voice content. In particular, it can capture and record the content written and drawn by the meeting presenter on the interaction device. Therefore, the generated meeting records can reflect all activities of the meeting in more detail. Furthermore, the intelligent meeting support system and the method for generating meeting records provided by the present invention have the ability to integrate voice content, text content, and chart content to generate comprehensive meeting records, and can correct and supplement the recorded incorrect and incomplete content through natural language processing and machine learning models to generate higher-quality meeting records.

[0064] The content disclosed above is only a preferred embodiment of the present invention and does not limit the scope of the claims of the present invention. Therefore, all equivalent technical modifications made based on the content of the specification and the attached drawings of the present invention shall be included in the scope of the claims of the present invention.

Explanation of Reference Numerals

[0065] 1 Intelligent Conference Support System 11 Image Capture Device 12 Image Analysis Device 121 Optical Character Recognition Circuit 122 Chart Recognition Circuit 123 Motion Display Recognition Circuit 13 Voice Input Device 14 Voice Analysis Device 141 Voice Recognition Circuit 142 Voice Signature Identification Circuit 15 Clock Circuit 16 Artificial Intelligence Processing Circuit 17 Output Device M1, M2, M3 Meeting Records 2 Dialogue Device 3 Meeting Environment 4 Image 5 Meeting Presenter 1211, 1221 Image Selection Circuit 1212, 1222, 1231 Image Segmentation Circuit 1213 Image Reconstruction Circuit 1214 Optical Character Recognition Engine 1215, 1224, 1233, 1415 Timestamp-Event Synthesis Circuit 1223 Chart Recognition Engine 12231 Circle Chart Recognition Engine 12232 Line Chart Recognition Engine 12233 Bar Chart Recognition Engine 1232 Motion Display Recognition Engine 1411 Voice Preprocessing Circuit 1412 Voice Recognition Engine 1413 Cloud Recognition Unit 1414 Local Recognition Unit Sent1, Sent2, Sent3, Sent4, Sent5 Texts Sp1, Sp2 Speakers St1, St2, St3, St4, St5 Timestamps QS Quick Summary MS1 Title MS2 Summary Content MS3 Keywords MS4 Pain Points MS5 Operation Items MS6 Chart B1 Checkbox B2, B3, B4, B5 Buttons S11, S12, S121, S122, S123, S124, S13, S14, S15 Steps

Claims

1. An image capture device arranged to capture an image displayed by a dialogue device during a meeting, and An image analysis device communicably connected to the image capture device, which executes an image analysis process on the image and generates a first meeting record recording the image content. An intelligent meeting support system, characterized by comprising the above.

2. The intelligent meeting support system according to claim 1, wherein the image content includes text content and / or chart content in the image, and the image analysis process includes a text recognition process and a chart recognition process.

3. The image analysis device includes An optical character recognition circuit arranged to execute the text recognition process on the image so as to generate a first event record recording the text content, and A chart recognition circuit arranged to execute the chart recognition process on the image so as to generate a second event record recording the chart content. The intelligent meeting support system according to claim 2, characterized by comprising the above.

4. The intelligent meeting support system according to claim 3, wherein the text content includes printed text content and / or handwritten text content in the image, and the optical character recognition circuit executes the text recognition process on the image to identify the printed text content and / or the handwritten text content in the image.

5. The intelligent meeting support system according to claim 3, wherein the chart content includes pie chart content, line chart content and / or bar chart content in the image, and the chart recognition circuit executes the chart recognition process on the image so as to identify the pie chart content, the line chart content and / or the bar chart content in the image.

6. The intelligent meeting support system according to claim 3, wherein the image analysis process further includes a motion display recognition process, and the image content includes the text content, the chart content and / or motion display content in the image.

7. The image analysis device further comprises a motion display recognition circuit arranged to execute the motion display recognition program on the image so as to generate a third event record recording the motion display content. The intelligent conference support system according to claim 6, wherein the image analysis device is configured to generate the first meeting record for recording the image content based on the first event record, the second event record, and the third event record.

8. A voice input device configured to generate voice data during the meeting period, A voice analysis device connected to the voice input device and executing a voice analysis program on the voice data to generate a second meeting record for recording a plurality of voice contents. The intelligent conference support system according to claim 7, comprising:

9. The intelligent conference support system according to claim 8, wherein the image capture device is configured to sequentially capture the plurality of images displayed on the interactive device during the meeting period, and the image analysis device generates the first meeting record for recording the plurality of image contents respectively.

10. The intelligent conference support system according to claim 9, wherein the image analysis device is configured to add a time stamp to each image content recorded in the first meeting record, and the voice analysis device is configured to add a time stamp to each voice content recorded in the second meeting record.

11. The intelligent conference support system according to claim 10, comprising an artificial intelligence processing circuit coupled to the image analysis device and the voice analysis device, receiving the first meeting record and the second meeting record, integrating the first meeting record and the second meeting record, and inputting the integrated record into a natural language processing and machine learning model for analysis to generate a third meeting record.

12. A method for generating a meeting record, comprising: Installing an image capture device to capture images displayed on an interactive device during a meeting period; Configuring an image analysis device to execute an image analysis program on the images to generate a first meeting record for recording image content.

13. The method according to claim 12, wherein the image content includes text content and / or chart content in the image, and the image analysis program includes a text recognition program and a chart recognition program.

14. Installing the image analysis device to execute the image analysis program on the images. An optical character recognition circuit is arranged to execute the text recognition program on the image and generate a first event record for recording the text content. The method according to claim 13, wherein a chart recognition circuit is set to execute the chart recognition program on the image and generate a second event record for recording the chart content.

15. The image analysis program further includes a motion display recognition program. The method according to claim 14, wherein the image content includes the text content, the chart content, and / or the motion display content on the image.

16. The step of setting the image analysis device to execute the image analysis program on the image further includes: Installing a motion display recognition circuit to execute the motion display recognition program on the image to generate a third event record for recording the motion display content. The method according to claim 15, wherein the image analysis device is set to generate a first meeting record for recording the image content based on the first event record, the second event record, and / or the third event record.

17. Installing an audio input device to generate audio data during the meeting. The method according to claim 16, wherein an audio analysis device is set to execute an audio analysis program on the audio data and generate a second meeting record for recording a plurality of audio contents.

18. The method according to claim 17, wherein the image capture device is configured to sequentially capture a plurality of the images displayed on the dialogue device during the meeting, and the image analysis device is configured to generate a first meeting record for recording each of the plurality of image contents.

19. The method according to claim 18, wherein the image analysis device is arranged to add a time stamp to each of the image contents recorded in the first meeting record, and the audio analysis device is arranged to add the time stamp to each of the audio contents recorded in the second meeting record.

20. The method according to claim 19, wherein an artificial intelligence processing circuit is set to integrate the first meeting record and the second meeting record, input them into a natural language processing and machine learning model, analyze them, and generate a third meeting record.

Images