CN115440233A - Bird singing analysis system based on deep learning model - Google Patents
Bird singing analysis system based on deep learning model Download PDFInfo
- Publication number
- CN115440233A CN115440233A CN202211061881.9A CN202211061881A CN115440233A CN 115440233 A CN115440233 A CN 115440233A CN 202211061881 A CN202211061881 A CN 202211061881A CN 115440233 A CN115440233 A CN 115440233A
- Authority
- CN
- China
- Prior art keywords
- data
- audio
- unit
- bird
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 41
- 238000013136 deep learning model Methods 0.000 title claims abstract description 20
- 230000005540 biological transmission Effects 0.000 claims abstract description 40
- 238000000034 method Methods 0.000 claims abstract description 6
- 238000012549 training Methods 0.000 claims description 24
- 238000000605 extraction Methods 0.000 claims description 21
- 238000004891 communication Methods 0.000 claims description 13
- 238000013135 deep learning Methods 0.000 claims description 11
- 239000004973 liquid crystal related substance Substances 0.000 claims description 10
- 238000004519 manufacturing process Methods 0.000 claims description 9
- 238000007781 pre-processing Methods 0.000 claims description 8
- 239000000284 extract Substances 0.000 claims description 7
- 238000001914 filtration Methods 0.000 claims description 7
- 230000003595 spectral effect Effects 0.000 claims description 7
- 238000013075 data extraction Methods 0.000 claims description 6
- 238000013500 data storage Methods 0.000 claims description 6
- 230000002708 enhancing effect Effects 0.000 claims description 6
- 238000003860 storage Methods 0.000 claims description 6
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 238000013144 data compression Methods 0.000 claims description 3
- 230000006837 decompression Effects 0.000 claims description 3
- 230000008030 elimination Effects 0.000 claims description 3
- 238000003379 elimination reaction Methods 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000011410 subtraction method Methods 0.000 claims description 3
- 238000002360 preparation method Methods 0.000 claims description 2
- 238000011160 research Methods 0.000 abstract description 6
- 238000001228 spectrum Methods 0.000 abstract description 5
- 230000006872 improvement Effects 0.000 description 10
- 241001465754 Metazoa Species 0.000 description 3
- 238000007906 compression Methods 0.000 description 3
- 230000006835 compression Effects 0.000 description 3
- 230000001965 increasing effect Effects 0.000 description 2
- 241000894007 species Species 0.000 description 2
- 241000251539 Vertebrata <Metazoa> Species 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000009395 breeding Methods 0.000 description 1
- 230000001488 breeding effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification techniques
- G10L17/26—Recognition of special voice characteristics, e.g. for use in lie detectors; Recognition of animal voices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/088—Non-supervised learning, e.g. competitive learning
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/03—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters
- G10L25/21—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters the extracted parameters being power information
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/03—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters
- G10L25/24—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters the extracted parameters being the cepstrum
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/27—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the analysis technique
- G10L25/30—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the analysis technique using neural networks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Signal Processing (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a bird singing analysis system based on a deep learning model, which comprises a data acquisition terminal, a data transmission terminal and a host control terminal, wherein the data acquisition terminal is connected with the data transmission terminal; the method and the device have the advantages that the Mel cepstrum coefficient characteristics and the energy spectrum density characteristics in the bird audio data are extracted and spliced, the identification performance is improved, meanwhile, the audio identification is applied to the bird identification based on a deep learning model, the collected bird singing audio data are preprocessed, the bird identification accuracy is improved, the error identification rate is greatly reduced, the industry admission standard of workers is reduced, the labor cost is reduced, meanwhile, the bird population scale prediction, the bird species number statistics in an area and the smooth implementation of protection planning work can be guaranteed through the analysis result, and effective help can be brought to the scientific research of birds.
Description
Technical Field
The invention relates to the technical field of bird singing analysis, in particular to a bird singing analysis system based on a deep learning model.
Background
Birds are the collective term for birds, higher vertebrates adapted to land and air life, birds being consumers in the food chain, an important component of the ecosystem. There are about 9000 kinds of birds existing in the world, 1186 kinds of birds exist in China, the number of bird species accounts for 48.6% of the total number of animal species in China and 13.1% of the total number of similar animal species in the world, the birds account for a very high proportion in the animal species in China, and the birds have very important functions in maintaining ecological balance and stability and have very high protection value and research value.
At present, the protection and scientific research of birds are still in the development stage, and China protects the birds by means of protecting the natural environment on which the birds live, establishing a natural protection area, artificially domesticating and breeding, salvation of the five precious birds and the like, and also scientifically researches the birds while protecting the birds.
At present, the traditional manual identification and statistical recording mode is still adopted for identifying birds in a certain area, and in the face of a large amount of workload and increasing labor cost, the identification and statistical mode gradually shows limitation, a person needs to be identified with higher professional bird knowledge, different characteristics of different kinds of birds are comprehensively known, manpower is consumed greatly, labor cost is higher, the birds are required to be in the range of vision of the person to be identified in the identification process, effective identification cannot be achieved for the birds flying or hidden in a tree crown, identification accuracy and identification efficiency are lower, and assistance cannot be brought to scientific research of the birds.
Disclosure of Invention
In view of the above problems, the invention aims to provide a bird singing analysis system based on a deep learning model, which solves the problems that the existing bird recognition mode has high requirements on recognition personnel, high labor cost, and low recognition accuracy and recognition efficiency.
In order to achieve the purpose of the invention, the invention is realized by the following technical scheme: the utility model provides a birds analytic system that sings based on degree of deep learning model, includes data acquisition terminal, data transmission terminal and host control terminal, the data acquisition terminal passes through data transmission terminal and host control terminal wireless connection, the data acquisition terminal is including the audio acquisition module that is used for gathering birds audio frequency of singing, be used for carrying out the data preprocessing module of preliminary treatment to gathering the audio frequency, be used for sending the data arrangement sending module of preliminary treatment audio data to data transmission terminal and be used for the power module of power supply, the data transmission terminal provides data transmission for data acquisition terminal and host control terminal, host control terminal is including the data receiving module that is used for receiving data transmission terminal transmission data, the birds audio data database that is used for saving different birds audio data, the data set preparation module that is used for making birds audio data set, the characteristic extraction module that is used for extracting birds audio data characteristic, the degree of deep learning module that is used for carrying out unsupervised training to the birds audio data characteristic of extraction, the contrastive analysis module and the liquid crystal display module that is used for showing contrastive analysis result and the keyboard input control system input control parameter.
The further improvement lies in that: the data acquisition terminal is installed on the tree crown through a support, the audio acquisition module comprises an audio acquisition device and a positioning chip, the audio acquisition device is used for acquiring audio played by birds, the positioning chip is used for providing position information of the data acquisition terminal, and the position information produced by the positioning chip is transmitted to the host control terminal through the data transmission terminal.
The further improvement is that: the data preprocessing module comprises an audio enhancing unit, a noise eliminating unit and an audio filtering unit, wherein the audio enhancing unit utilizes a DAC technology to enhance collected audio data, the noise eliminating unit utilizes a basic spectral subtraction method to preliminarily eliminate noise in audio, and the audio filtering unit utilizes a self-adaptive filter to filter the audio data subjected to preliminary noise elimination so as to further filter interference noise in the audio data.
The further improvement lies in that: the data arrangement sending module comprises a data arrangement unit for arranging the preprocessed audio data, a data compression unit for compressing the audio data and a data sending unit for sending the compressed audio data to the data transmission terminal, and the power supply module comprises a storage battery and a solar photovoltaic panel for charging the storage battery.
The further improvement lies in that: the data transmission terminal comprises a data transmission module for transmitting audio data, a data storage module for performing storage backup on the audio data and a wireless communication module, the data storage module performs body storage on the audio data and performs cloud backup on the audio data, and the wireless communication module comprises a Bluetooth communication unit for Bluetooth connection and a 5G network communication unit for 5G network connection.
The further improvement is that: the data receiving unit comprises a data receiving unit for receiving data transmitted by the data transmission terminal and a data decompression unit for decompressing the audio data, and the data set manufacturing module is used for carrying out data check and screening on the audio data of different varieties of birds in the bird audio database so as to manufacture a data set and mark bird variety labels corresponding to bird audio in the data set.
The further improvement lies in that: the feature extraction module comprises an audio input unit, a feature extraction unit and a feature splicing unit, the audio input unit inputs decompressed audio data into the feature extraction unit, the feature extraction unit extracts Mel cepstrum coefficient features and energy spectral density features of the audio data respectively, and the feature splicing unit splices the extracted Mel cepstrum coefficient features and the energy spectral density features.
The further improvement lies in that: the deep learning module comprises a pre-training unit and a fine-tuning unit, the pre-training unit carries out unsupervised training on spliced audio characteristic data by using a deep belief network model, and the fine-tuning unit further carries out global adjustment on weights from top to bottom by using a BP traditional neural network training method, so that the performance of the deep belief network model is optimal.
The further improvement is that: the unsupervised training comprises the following specific steps: feature extraction is introduced into training data to obtain an input layer, and then an output layer is calculated by adopting a CD algorithm according to a feature vector of the input layer, so that iterative training is performed.
The further improvement lies in that: the contrastive analysis module comprises a data extraction unit, a data comparison unit and a result output unit, the data extraction unit extracts audio characteristic data which are trained in the deep learning module, the data comparison unit performs contrastive analysis on the extracted audio characteristic data and bird audio data in the bird audio data set, bird varieties corresponding to the audio data are confirmed, and the result output unit outputs contrastive analysis results to the liquid crystal display module to display the contrastive analysis results.
The invention has the beneficial effects that: the method and the device have the advantages that the Mel cepstrum coefficient characteristics and the energy spectrum density characteristics in the bird audio data are extracted and spliced, the identification performance is improved, meanwhile, the audio identification is applied to the bird identification based on a deep learning model, the collected bird singing audio data are preprocessed, the bird identification accuracy is improved, the error identification rate is greatly reduced, the industry admission standard of workers is reduced, the labor cost is reduced, meanwhile, the bird population scale prediction, the bird species number statistics in an area and the smooth implementation of protection planning work can be guaranteed through the analysis result, and effective help can be brought to the scientific research of birds.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a schematic diagram of the system architecture of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Example one
Referring to fig. 1, the embodiment provides a birds singing analysis system based on a deep learning model, which comprises a data acquisition terminal and a host control terminal, wherein the data acquisition terminal is wirelessly connected with the host control terminal through a data transmission terminal and performs data transmission, the data acquisition terminal comprises an audio acquisition module, a data preprocessing module, a data sorting and sending module and a power supply module, the audio acquisition module acquires bird singing audio data, the data preprocessing module preprocesses the acquired audio data to make the acquired audio data clearer, the subsequent comparison and analysis are facilitated, the data sorting and sending module sends the preprocessed audio data to the data transmission terminal, the power supply module supplies power to the data acquisition terminal, the host control terminal comprises a data receiving module, a bird audio database, a data set manufacturing module, a feature extraction module, a deep learning module, a comparison and analysis module and a liquid crystal display module, the data receiving module receives the data transmitted by the data transmission terminal, the bird audio database stores audio data of different birds, the data set manufacturing module manufactures bird audio data sets according to the bird audio database, the feature extraction module extracts features of the bird audio data, the audio data set extracted by the deep learning module performs comparison and analysis of a control keyboard comparison and display module performs comparison and a liquid crystal display control display module to the liquid crystal display keyboard.
Data acquisition terminal passes through the support mounting on the tree crown, is convenient for be close with birds, improves birds and sing audio frequency collection efficiency, and audio frequency collection module is including the audio frequency collector who is used for gathering birds and sing the audio frequency and the location chip that is used for providing data acquisition terminal positional information, and the positional information of location chip production passes through data transmission terminal transmission to host control terminal, the position at each data acquisition terminal of the audio-visual understanding of the user of being convenient for.
The data preprocessing module comprises an audio enhancing unit, a noise eliminating unit and an audio filtering unit, wherein the audio enhancing unit utilizes a DAC (digital-to-analog converter) technology to enhance collected audio data, the noise eliminating unit utilizes a basic spectral subtraction method to preliminarily eliminate noise in audio, the audio filtering unit utilizes an adaptive filter to carry out filtering processing on the audio data subjected to preliminary noise elimination, interference noise in the audio data is further filtered, preprocessing is carried out on the audio data through singing collected birds, and therefore the interference noise in the audio data can be taken out, the audio is clearer, and subsequent deep learning and contrastive analysis are facilitated.
Data arrangement sending module is including being used for carrying out the data arrangement unit of arranging in order to the audio data after the preliminary treatment, a data compression unit that is used for carrying out the compression to the audio data after the arrangement and a data sending unit that is used for sending the audio data after the compression to data transmission terminal, it is more high-efficient to make data transmission through the compression to data, thereby the operating efficiency of system has been improved, power module includes the battery and is used for the solar photovoltaic board that charges for the battery, through the solar energy power supply, avoid the problem that need frequently change the power, and energy-concerving and environment-protective.
The data transmission terminal comprises a data transmission module for transmitting audio data, a data storage module and a wireless communication module for storing and backing up the audio data, the data storage module stores the audio data and backs up the audio data in a cloud, the wireless communication module comprises a Bluetooth communication unit for providing Bluetooth connection and a 5G network communication unit for providing 5G network connection, and the data transmission terminal and the host control terminal are provided with two different data transmission channels, so that the system can be applicable to various communication environments and is convenient to use.
The data receiving unit comprises a data receiving unit for receiving data transmitted by the data transmission terminal and a data decompression unit for decompressing the audio data, and the data set manufacturing module is used for carrying out data check and screening on the audio data of different varieties of birds in the bird audio database so as to manufacture the data set and mark bird variety labels corresponding to bird audio in the data set, thereby facilitating the follow-up confirmation of comparison and analysis results.
The feature extraction module comprises an audio input unit, a feature extraction unit and a feature splicing unit, wherein the audio input unit inputs decompressed audio data into the feature extraction unit, the feature extraction unit extracts Mel cepstrum coefficient features and energy spectrum density features of the audio data respectively, the feature splicing unit splices the extracted Mel cepstrum coefficient features and the energy spectrum density features, and the extracted feature data are representative through extracting the Mel cepstrum coefficient features and the energy spectrum density features, so that the accuracy of an analysis result is improved.
The deep learning module comprises a pre-training unit and a fine-tuning unit, the pre-training unit carries out unsupervised training on spliced audio characteristic data by using a deep belief network model, and the fine-tuning unit further carries out global adjustment on weights from top to bottom by using a BP traditional neural network training method, so that the performance of the deep belief network model is optimal, and the training of the audio characteristic data is more efficient.
The specific steps of the unsupervised training are as follows: feature extraction is introduced into training data to obtain an input layer, and an output layer is calculated by adopting a CD algorithm according to the feature vector of the input layer, so that iterative training is performed.
The contrastive analysis module comprises a data extraction unit, a data contrast unit and a result output unit, the data extraction unit extracts audio characteristic data which are trained in the deep learning module, the data contrast unit contrasts and analyzes the extracted audio characteristic data and bird audio data in a bird audio data set, and confirms bird varieties corresponding to the audio data, the result output unit outputs contrastive analysis results to the liquid crystal display module for display, and a user can conveniently visually know bird singing analysis results.
During the use, utilize data acquisition terminal to gather the audio data that birds singed in the forest to audio data through data acquisition terminal with gathering send host control terminal to carry out deep study and contrastive analysis, and obtain the analysis result that birds singed
Example two
The embodiment provides a birds analytic system that sings based on degree of deep learning model, still include remote mobile terminal, remote mobile terminal is hand-carried by the user, remote mobile terminal and host control terminal wireless connection, remote control terminal includes smart mobile phone and panel computer, all be equipped with the APP that is used for control system on smart mobile phone and the panel computer, through setting up remote mobile terminal, make the user not also can realize the control to the system at host control terminal, the portability of system has been increased.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (10)
1. The utility model provides a birds singing analytic system based on degree of deep learning model which characterized in that: including data acquisition terminal, data transmission terminal and host control terminal, the data acquisition terminal passes through data transmission terminal and host control terminal wireless connection, the data acquisition terminal is including the audio acquisition module that is used for gathering birds to sing the audio frequency, be used for carrying out the data preprocessing module of preliminary treatment to gathering the audio frequency, be used for with the data arrangement sending module of preliminary treatment audio data transmission to data transmission terminal and be used for the power module of power supply, the data transmission terminal provides data transmission for data acquisition terminal and host control terminal, host control terminal is including the data receiving module that is used for receiving data transmission terminal transmission data, the birds audio database that is used for saving different birds audio data, the data set preparation module that is used for making birds audio data set, the feature extraction module that is used for extracting birds audio data feature, be used for carrying out the deep learning module of unsupervised training to the birds audio data feature of extraction, be used for carrying out contrastive analysis module and the liquid crystal display module to birds audio data feature and be used for carrying out contrastive analysis to birds audio data feature, the liquid crystal display module is including the liquid crystal display that is used for showing contrastive analysis result and the touch keyboard that is used for system input control parameter.
2. The bird singing analysis system based on the deep learning model of claim 1, wherein: the data acquisition terminal is installed on the tree crown through a support, the audio acquisition module comprises an audio acquisition device and a positioning chip, the audio acquisition device is used for acquiring audio played by birds, the positioning chip is used for providing position information of the data acquisition terminal, and the position information produced by the positioning chip is transmitted to the host control terminal through the data transmission terminal.
3. The bird singing analysis system based on the deep learning model according to claim 1, characterized in that: the data preprocessing module comprises an audio enhancing unit, a noise eliminating unit and an audio filtering unit, wherein the audio enhancing unit utilizes a DAC technology to enhance collected audio data, the noise eliminating unit utilizes a basic spectral subtraction method to preliminarily eliminate noise in audio, and the audio filtering unit utilizes a self-adaptive filter to filter the audio data subjected to preliminary noise elimination so as to further filter interference noise in the audio data.
4. The bird singing analysis system based on the deep learning model according to claim 1, characterized in that: the data arrangement sending module comprises a data arrangement unit for arranging the preprocessed audio data, a data compression unit for compressing the audio data and a data sending unit for sending the compressed audio data to the data transmission terminal, and the power supply module comprises a storage battery and a solar photovoltaic panel for charging the storage battery.
5. The bird singing analysis system based on the deep learning model of claim 1, wherein: the data transmission terminal comprises a data transmission module for transmitting audio data, a data storage module for storing and backing up the audio data and a wireless communication module, the data storage module stores the audio data and carries out cloud backup, and the wireless communication module comprises a Bluetooth communication unit for Bluetooth connection and a 5G network communication unit for 5G network connection.
6. The bird singing analysis system based on the deep learning model of claim 1, wherein: the data receiving unit comprises a data receiving unit for receiving data transmitted by the data transmission terminal and a data decompression unit for decompressing the audio data, and the data set manufacturing module is used for carrying out data check and screening on the audio data of different varieties of birds in the bird audio database so as to manufacture a data set and mark bird variety labels corresponding to bird audio in the data set.
7. The bird singing analysis system based on the deep learning model of claim 1, wherein: the feature extraction module comprises an audio input unit, a feature extraction unit and a feature splicing unit, wherein the audio input unit inputs decompressed audio data into the feature extraction unit, the feature extraction unit respectively extracts Mel cepstrum coefficient features and energy spectral density features of the audio data, and the feature splicing unit splices the extracted Mel cepstrum coefficient features and the energy spectral density features.
8. The bird singing analysis system based on the deep learning model of claim 7, wherein: the deep learning module comprises a pre-training unit and a fine-tuning unit, the pre-training unit performs unsupervised training on spliced audio characteristic data by using a deep belief network model, and the fine-tuning unit performs further global adjustment on weights from top to bottom by using a BP traditional neural network training method, so that the performance of the deep belief network model is optimal.
9. The bird singing analysis system based on the deep learning model of claim 8, characterized in that: the unsupervised training comprises the following specific steps: feature extraction is introduced into training data to obtain an input layer, and an output layer is calculated by adopting a CD algorithm according to the feature vector of the input layer, so that iterative training is performed.
10. The bird singing analysis system based on the deep learning model according to claim 1, characterized in that: the comparison and analysis module comprises a data extraction unit, a data comparison unit and a result output unit, the data extraction unit extracts audio characteristic data which are trained in the deep learning module, the data comparison unit performs comparison and analysis on the extracted audio characteristic data and bird audio data in a bird audio data set, bird varieties corresponding to the audio data are confirmed, and the result output unit outputs comparison and analysis results to the liquid crystal display module for display.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211061881.9A CN115440233A (en) | 2022-09-01 | 2022-09-01 | Bird singing analysis system based on deep learning model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211061881.9A CN115440233A (en) | 2022-09-01 | 2022-09-01 | Bird singing analysis system based on deep learning model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115440233A true CN115440233A (en) | 2022-12-06 |
Family
ID=84243673
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211061881.9A Pending CN115440233A (en) | 2022-09-01 | 2022-09-01 | Bird singing analysis system based on deep learning model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115440233A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117612537A (en) * | 2023-11-27 | 2024-02-27 | 北京林业大学 | Bird song intelligent monitoring system based on cloud limit cooperative control |
-
2022
- 2022-09-01 CN CN202211061881.9A patent/CN115440233A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117612537A (en) * | 2023-11-27 | 2024-02-27 | 北京林业大学 | Bird song intelligent monitoring system based on cloud limit cooperative control |
| CN117612537B (en) * | 2023-11-27 | 2024-06-07 | 北京林业大学 | Bird song intelligent monitoring system based on cloud limit cooperative control |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3435374A1 (en) | Method and device for voice data processing and storage medium | |
| CN104700829B (en) | Animal sounds Emotion identification system and method | |
| CN107103903A (en) | Acoustic training model method, device and storage medium based on artificial intelligence | |
| CN115440233A (en) | Bird singing analysis system based on deep learning model | |
| CN114027052A (en) | Illumination regulation and control system for plant reproductive development | |
| CN105919591A (en) | Surface myoelectrical signal based sign language recognition vocal system and method | |
| CN107799114A (en) | A kind of pig cough sound recognition methods and system | |
| CN107506434A (en) | Method and apparatus based on artificial intelligence classification phonetic entry text | |
| CN110059765B (en) | Intelligent mineral identification and classification system and method | |
| CN110110095A (en) | A kind of power command text matching technique based on shot and long term memory Recognition with Recurrent Neural Network | |
| CN111145726B (en) | Deep learning-based sound scene classification method, system, device and storage medium | |
| CN106340299A (en) | Speaker recognition system and method in complex environment | |
| CN109448700A (en) | A kind of built-in speech recognition system and method based on improvement MFCC algorithm | |
| CN114791771A (en) | Interaction management system and method for intelligent voice mouse | |
| CN110599232A (en) | Consumption group analysis method based on big data | |
| Zhang et al. | Research on spectrum sensing system based on composite neural network | |
| CN116884416A (en) | Wild animal audio acquisition and detection system, method, storage medium and electronic equipment | |
| CN116600452A (en) | Color regulation and control method and system based on intelligent atmosphere lamp strip | |
| Zhang et al. | A novel insect sound recognition algorithm based on MFCC and CNN | |
| CN116013330A (en) | Forest area birds biological investigation system | |
| CN114764575B (en) | Multi-modal data classification method based on deep learning and time sequence attention mechanism | |
| CN115115880A (en) | Mushroom quality grading method based on deep learning | |
| Bashit | A comprehensive solar powered remote monitoring and identification of Houston Toad call automatic recognizing device system design | |
| CN112992156B (en) | Power distribution network dispatching identity authentication system based on voiceprint authentication | |
| CN115481674A (en) | Single cell type intelligent identification method based on deep learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |