Intelligent voice scoring system based on voice recognition and automatic event marking method

Abstract

The invention discloses an intelligent voice scoring system based on voice recognition and an automatic event marking method, and relates to the technical field of intelligent electronic equipment. According to the method, a score indicator host and a plurality of wireless microphone modules (slave machines) are included; a voice recognition and intelligent semantic understanding system, a multi-wireless microphone collaborative input system, a personalized name management system, a scoring and data management system, a score and main event recording and broadcasting system, a personal account data synchronization system, a data backup and remote upgrading system, a power-off protection system and a competition rule management system are deployed. The invention discloses a switchable local computing and cloud computing co-processing architecture and system. According to the method, real-time scoring is carried out through voice instructions, main data in the competition process are recorded, revocation and modification operations are supported, and a statistical report is automatically generated after the competition is finished. The problems that traditional scoring efficiency is low, interactivity is poor, multi-language recognition is poor, simultaneous participation of multiple persons cannot be met, data management and tracing functions are lacked, and the scoring function is single are solved, and the system is suitable for scenes such as sports competitions and chess and card competitions.

Description

Technical Field

[0001] This invention relates to the field of intelligent electronic device technology, specifically to an intelligent voice scoring system based on speech recognition and an automatic event tagging method. Background Technology

[0002] Traditional competition scoring systems rely on manual recording or simple electronic scoring devices, which have the following drawbacks: 1. Cumbersome and error-prone operation: Traditional scoring systems rely on manual recording or manual (remote) input, which is inefficient and prone to errors, especially when multiple people participate, making it difficult to guarantee the accuracy and real-time nature of scores. 2. Poor interactivity: Most traditional devices do not provide voice interaction functions, making it impossible to conveniently update scores and provide feedback through voice commands, thus reducing the user experience. 3. Inability to accommodate multiple participants simultaneously: In multi-person competitions, existing devices cannot effectively handle multiple user inputs, leading to input failures and delayed score updates. 4. Lack of data management and traceability functions: Traditional scorers lack automatic recording, statistical, and export functions, making it difficult to conveniently view and analyze historical data. 5. Limited scoring function: Data recording only stores the final score, lacking process traceability and statistical analysis functions, and also lacking the ability to record important events during the competition.

[0003] Therefore, there is an urgent need for an intelligent voice scoring system and an automatic event tagging method that can automatically process scores, support multi-user voice interaction, and have data management functions, in order to improve scoring efficiency and user experience. Summary of the Invention

[0004] Purpose of the invention: This invention aims to provide an intelligent voice scoring system and an automatic event tagging method based on speech recognition. By integrating multilingual recognition, intelligent semantic error correction, personalized name setting, and deep data analysis functions, it solves the problems of low efficiency, insufficient accuracy, and incomplete process recording in existing scoring methods, thereby effectively improving scoring efficiency, ensuring scoring accuracy, and guaranteeing the integrity of the scoring process. Technical solution

[0005] The present invention provides an intelligent voice scoring system and automatic event tagging method based on speech recognition, comprising the following hardware devices and software systems:

[0006] Hardware devices: ① Main Unit: Includes a touchscreen, processor, storage unit, battery, wireless communication module, power interface, display interface (supports HDMI, etc.), and internet connectivity (WiFi, 4G, etc.). The screen features high brightness display, automatic brightness adjustment, glove mode, and waterproof design. ② Multiple Wireless Microphone Modules (Slave Units): Features noise reduction and sensitivity adjustment functions, and connects to the main unit via the wireless communication module.

[0007] Software system: ① Speech Recognition and Intelligent Semantic Understanding System: This system uses speech-to-text (STT) technology, supporting speech recognition in multiple languages ​​including Mandarin and local dialects; it integrates a Large Language Model (LLM) to perform semantic error correction on the STT output text (e.g., correcting "Red Team +1 point" to "Red Team +1 point"), and parses user commands (e.g., "Zhang San +2 points"); it supports user-defined dialect sample training and can also remotely upgrade language packs via OTA. ② Multi-Wireless Microphone Collaborative Input System: This system supports multiple wireless microphone devices connecting to the host via wireless communication protocols such as Bluetooth, Wi-Fi, and 2.4G; it has built-in anti-interference algorithms (e.g., signal filtering, frequency switching) and an automatic reconnection mechanism; it adds a priority recognition mechanism (e.g., referee microphones have higher priority than ordinary players) and keyword filtering (pre-identifying commands containing keywords such as "add," "deduct," and "cancel"). ③ Personalized Name Management System: Users can customize their team name (e.g., "Zhang San Team," "Li Si Team"), with default names in red, yellow, blue, and green (short syllables, clear pronunciation, reducing misidentification rate); supports name voice verification (the system confirms the name by repeating it after the user sets it) and fuzzy matching (e.g., automatically matching "Zhang San" when the user says "Zhang Shan"). ④ Scoring and Data Management System: Records scores and major events in real time (timestamp, operation type, score before / after modification, key event tags), supports voice undo (requires user confirmation to "undo the previous step") and manual modification via touch screen (supports operation log back); data is stored encrypted and supports exporting to Excel / CSV format (users can customize statistical dimensions: scoring rounds, key events, player names, etc.). ⑤ Score and Major Event Recording and Broadcasting System: After each score change or major event (including but not limited to reward reasons or penalty information, supporting the marking of key scoring points or points lost, etc.) update, the system automatically broadcasts the current score and related statistics via voice, improving on-site interactivity and transparency. ⑥ Personal Account Data Synchronization System: Supports synchronizing match records to personal accounts via QR code scanning or other methods, and managing statistics according to a timeline. ⑦ Data Backup and Remote Upgrade System: The system has 4G and WiFi communication capabilities, enabling remote synchronization, data backup, and system upgrades via the internet. ⑧ Power Outage Protection System: Features cloud storage and power outage protection mechanisms, preserving data even offline or during power outages, and ensuring uninterrupted operation after recovery. ⑨ Match Rule Management System: Allows setting match time and score limits, with automatic voice prompts (e.g., "Match time is up") when the time / score targets are met; supports countdown, timer pause (e.g., pausing the timer when the match is interrupted), and time synchronization (synchronizing NTP time when connected to the internet, and using a high-precision clock module when offline). It also supports multiple scoring modes, including bonus mode and battle mode, allowing for flexible configuration of bonus and penalty rules to adapt to different game or competition scenarios.⑩ Switchable local and cloud computing collaborative processing architecture and system: The local processor has noise filtering, keyword extraction, person mapping, basic speech recognition and scoring functions, and can independently complete recognition and scoring in offline or weak network environments; the cloud processing module uses cloud computing resources or large-scale artificial intelligence models to perform high-precision speech recognition, multilingual semantic analysis and complex competition logic judgment. The system can dynamically select local mode, cloud mode or hybrid mode according to network conditions and task requirements.

[0008] Beneficial Effects: Compared with existing technologies, this invention has the following significant advancements: ① Convenient Operation: Score recording is achieved through voice interaction, eliminating the need for manual operation and improving scoring efficiency. ② Accurate Recognition Enhances User Experience: Combining a large model and STT technology, it supports multilingual and dialect recognition, tolerates speech-to-text errors, and further improves recognition accuracy after networking. This enhances system fault tolerance and user experience. ③ Multi-Scenario Adaptability: Supports multiple participants, offers flexible wireless connectivity, and has offline and outdoor operation capabilities. Network connectivity expands online application scenarios, making it suitable for various competitions and entertainment scenarios. ④ Rich Functionality: It features score recording, broadcasting, display, undoing, modification, data export, and statistical analysis functions. After networking, it adds online synchronization and live streaming capabilities to meet professional needs. ⑤ Improved Data Integrity and Value: It enables complete traceability of the competition process, and custom dimension export (Excel / CSV) supports rapid verification, solving the deficiency of traditional devices that only store the final score. The newly added key event recording and statistical functions enhance the depth and value of competition data, which can be used to select the best scorer and analyze competition trends. ⑥ Enhanced Competition Standardization and Professionalism: Time and score limits combined with automatic voice prompts make the competition more standardized (experimental data: network-synchronized NTP test), avoiding timeouts or early termination and improving the professionalism of the event. ⑦ Strong Expandability: Supports expansion in wireless communication methods, number of participants, and languages; possesses OTA upgrade capabilities; and network connectivity allows for future integration of more third-party services and feature upgrades, facilitating subsequent feature upgrades and optimizations. Detailed Implementation

[0009] This invention discloses a multi-user intelligent scoring system based on speech recognition. Through a modular structure design, it enables accurate and efficient real-time score management and key event statistics in multi-player competition scenarios. The following describes the configuration, working principle, communication method, and functional implementation of each module:

[0010] 1. Device structure and configuration of main components The system includes a main unit and multiple wireless microphone modules. The main unit is responsible for overall control, data processing, and communication with the cloud server; the wireless microphone modules are responsible for capturing user voice and displaying the results. The specific configuration is as follows: ① Main Unit: Composed of a screen, motherboard, battery, speaker, external expansion magnetic interface, wireless power receiving module, and main unit detection sensors. The motherboard integrates a processor, power management, voice recognition, local computing unit, wireless communication module (Bluetooth, WiFi, 4G), and storage module. ② Wireless Microphone Module: Several modules are available, each slave unit equipped with an independent motherboard, microphone, display screen, battery, wireless communication module, magnetic interface, and wireless power receiving module. Used to recognize the voice of a designated user and send the collected data to the main unit.

[0011] like Figure 1As shown: ① Screen: A touch control screen used to display score data, operating interface, and device mode selection. It also supports manual input and score adjustment. ② Motherboard: Integrates a main control processor, capable of driving the screen, receiving microphone input, and controlling speaker output. It supports Bluetooth or 2.4G wireless communication with up to four slave devices. It also features battery management, 4G communication, WiFi connection, external expansion device connection, power supply control, button input, volume adjustment, and local algorithm operation. ③ Main unit battery: Provides independent power to the main unit, supporting battery life and mobile usage needs. ④ Speaker: Used for voice broadcasting of score information or system interaction prompts, realizing sound information output. ⑤ External expansion magnetic interface: The external expansion magnetic interface is used to achieve physical connection and functional integration with other devices, enabling the device to be attached to or be attached to external devices. ⑥: Main unit wireless power receiving module: can receive power from other external devices (such as wireless charging docks or expansion modules) to realize the wireless charging function of the main unit; ⑦: Main unit wireless power supply module: provides wireless power transmission to the slave modules attached to the main unit, enabling the slave modules to provide power supply support when there is no independent power source; ⑧: Main unit external expansion magnetic interface: used to detect the connection status of the slave devices, whether the magnetic connection is aligned, and control whether the wireless power supply module is working; it can also detect whether the expansion host is connected and switch to multi-device collaborative linkage mode accordingly; ⑨: Main unit detection sensor: detects the status of the slave devices, controls the operation of the wireless power supply module, detects whether the expansion host is matched, and selects multi-device linkage mode. ⑩: Slave Battery: Provides independent power to the slave device, maintaining normal operation in the absence of wireless power supply; ⑪: Slave Motherboard: Capable of driving the local screen, receiving microphone voice input, conducting Bluetooth / 2.4G wireless communication with the host, and managing battery power; ⑫: Matches the host's magnetic interface for physical magnetic connection between the slave module and the host; ⑬: Slave Detection Sensor: Detects the status of external devices, used to trigger power supply control, device identification, and system configuration judgment, and can also be extended to control external display devices; ⑭: Slave Wireless Power Module: Receives wireless power from the host to enable wireless charging for the slave, ensuring stable operation of the slave over long periods. ③ User Terminal Equipment: The system can connect to user-worn wireless microphones, display devices, etc., for listening to voice broadcasts and inputting voice commands. ④ Cloud Server: Provides the system with advanced speech recognition, multi-semantic analysis, large model inference, and data storage services.

[0012] 2. Working principle and process like Figure 2As shown. ① Voice Input and Preprocessing: Users input voice commands (such as "Player A scores 3 points" or "Player B steals once") through the microphone of the slave device or a personal terminal. The slave device sends the voice data to the host device via wireless communication. The host device performs noise filtering, keyword extraction, and preliminary speech recognition. ② Dynamic Calculation Mode Selection: Depending on the network status and task type, the system can select local processing mode, cloud processing mode, or hybrid mode. When the network is stable, the host device uploads voice data to the cloud server to perform high-precision recognition and complex semantic analysis using a large model; when the network is unstable or offline, the host device uses a local model to complete basic recognition and scoring, and synchronizes the data after the network is restored. ③ Score Update and Event Recording: After parsing the recognition results, the system automatically updates the score of the corresponding user and records key events, such as shots, steals, fouls, etc.; after each update, it is broadcast through the host device's speaker or the user terminal, and the real-time score and key statistics are displayed on the slave device and host device screens. ④ Error Correction and Adjustment: Users can undo or modify the previous record by inputting error correction commands through voice commands or touch screens. The system updates the score and statistics after confirmation. ⑤ Data Statistics and Analysis: The system continuously aggregates scores and event information in the background, automatically generating statistical reports for each game or the entire match, including each user's score, number of shots, shooting percentage, number of steals, number of fouls, etc., and can select MVP or key scorers based on algorithms. Users can view and export statistical reports by scanning a QR code or logging in with their account.

[0013] 3. System expansion and multi-mode support ① Supports networking of multiple host and slave devices; multiple hosts can be connected via wireless network, making the system suitable for large-scale competitions. ② Scoring modes can be configured with bonus scoring mode, battle mode, and custom mode containing various logics.

[0014] 4. Modular design and maintainability ① The master and slave units use a magnetic interface and wireless power supply for easy installation and removal. ② The system is designed with an open architecture, allowing for the addition of new slave modules, sensors, or external data interfaces to expand functionality. Attached Figure Description Figure 1 : A schematic diagram of the host and slave modules, labeled with components such as screen, motherboard, battery, speaker, magnetic interface, wireless power supply module, and detection sensors; Figure 2 The system topology diagram and scoring and event logging flowchart illustrate the communication connections and data flow between the host, the user's wireless microphone (slave) or terminal device, and the remote server; and describe the processes of voice input, preprocessing, recognition, score updating, key event logging, and data analysis.

Claims

1. A speech scoring system based on speech recognition and an automatic event marking method, characterized in that, Including hardware devices and software systems: The hardware device includes: a scoring unit host, comprising a screen, processor, storage unit, battery, wireless communication module, power interface, and display interface, with internet connectivity; the screen features high brightness display, automatic brightness adjustment, glove mode, and waterproof design. Multiple wireless microphone modules with noise reduction and sensitivity adjustment functions can connect to the host via the wireless communication module. The software system includes: a speech recognition and intelligent semantic understanding system, employing speech-to-text technology, supporting multilingual speech recognition, accessing a large language model to perform semantic error correction on the speech-to-text output and parsing user commands, supporting user-defined dialect sample training, and allowing for remote language pack upgrades via OTA. The system includes a multi-microphone collaborative input system that supports multiple wireless microphone devices connecting to the host via wireless communication protocols. It features built-in anti-interference algorithms and an automatic reconnection mechanism, along with priority recognition and keyword filtering. A personalized name management system allows users to customize their competition names; the default name is a clearly pronounced short syllable, and it supports voice verification and fuzzy matching. A scoring and data management system records scores and key events in real time, supporting voice cancellation and manual modification via touchscreen. Data is stored encrypted and can be exported to Excel / CSV formats; users can customize statistical dimensions. A score and key event recording and broadcasting system automatically announces scores and key events via voice after each score change or key event update. The system features: broadcasting current scores and related statistics; a personal account data synchronization system that supports synchronizing match records to personal accounts and managing statistics according to a timeline; a data backup and remote upgrade system with 4G and WiFi communication capabilities, enabling remote synchronization, data backup, and system upgrades via the internet; a power outage protection system with cloud storage and power outage protection mechanisms, ensuring data retention even offline or during power outages and uninterrupted operation upon recovery; a match rule management system that allows setting match time and score limits, providing automatic voice prompts when time / score targets are met, supporting countdown, timer pause, and timer synchronization, multiple scoring modes, and customizable point deduction rules; and a switchable local and cloud computing collaborative processing architecture and system. The local processor features noise filtering, keyword extraction, person mapping, basic speech recognition, and scoring functions, enabling independent recognition and scoring in offline or weak network environments. The cloud processing module utilizes cloud computing resources or large-scale artificial intelligence models to perform high-precision speech recognition, multilingual semantic analysis, and complex match logic judgments. The system can dynamically select local mode, cloud mode, or hybrid mode based on network conditions and task requirements.

2. The intelligent voice scoring system based on speech recognition according to claim 1, characterized in that, The host's motherboard integrates a main control processor, which has the ability to drive the screen, receive microphone input, and control speaker output. It supports Bluetooth or 2.4G wireless communication with multiple slave devices, and also has battery management, 4G communication, WiFi connection, external expansion device connection, power supply control, button input, volume adjustment, and local algorithm operation functions.

3. The intelligent voice scoring system based on speech recognition according to claim 1, characterized in that, The wireless microphone module is equipped with an independent motherboard, microphone, display screen, battery, wireless communication module, magnetic interface and wireless power receiving module, used to recognize the voice of a designated user and send the collected data to the host.

4. The intelligent voice scoring system based on speech recognition according to claim 1, characterized in that, The system can be connected to a user's wireless microphone and display device for listening to voice broadcasts and inputting voice commands.

5. The intelligent voice scoring system based on speech recognition according to claim 1, characterized in that, This includes cloud servers that provide the system with advanced speech recognition, multi-semantic analysis, large-scale model inference, and data storage services.

6. An automatic event marking method based on the intelligent voice scoring system according to any one of claims 1-5, characterized in that, Includes the following steps: Voice Input and Preprocessing: Users input voice commands through the microphone of the wireless microphone module or their personal terminal. The wireless microphone module transmits the voice data to the host via wireless communication. The host performs noise filtering, keyword extraction, and preliminary speech recognition. Dynamic Calculation Mode Selection: Depending on network status and task type, the system selects local processing mode, cloud processing mode, or a hybrid mode. When the network is stable, the host uploads voice data to the cloud server to perform high-precision recognition and complex semantic analysis using a large model. When the network is unstable or offline, the host uses a local model to complete basic recognition and scoring, and synchronizes the data after the network is restored. Score Update and Event Recording: After parsing the recognition results, the system automatically updates the corresponding user's score and records key events. After each update, the system broadcasts the results via the host's speaker or the user terminal, and displays the real-time score and key statistics on the wireless microphone module and the host screen. Error Correction and Adjustment: Users can undo or modify the previous record by inputting error correction commands via voice or touchscreen. The system updates the score and statistics after confirmation. Data statistics and analysis: The system continuously summarizes scores and event information in the background and automatically generates statistical reports for each game or the entire game. Users can view and export statistical reports by scanning a QR code or logging in with their account.

7. The automatic event tagging method according to claim 6, characterized in that, It supports networking of multiple hosts and slaves, and can connect multiple hosts via wireless network for use in large-scale events.

8. The automatic event tagging method according to claim 6, characterized in that, The scoring mode can be configured with a bonus mode, a battle mode, and a custom mode containing multiple logics.

Images