What is Image Recognition and How it is Used?

Image recognition refers to technologies that identify places, logos, people, objects, buildings, and several other variables in digital ... AllCourses DataScienceCourses Master’sinDataScience M.TechinDataScience(Bangalore) DataScienceSpecialization PGinBusinessAnalytics DecisionMakingUsingDataScience DataScienceandMachineLearning ArtificialIntelligenceCourses PGP–ArtificialIntelligenceandMachineLearning(Online) PGPinArtificialIntelligenceandMachineLearning(Classroom) PGP–ArtificialIntelligenceforLeaders PGinMachineLearning PGDiplomainComputerScienceandAI–IIITDelhi AIforHighSchoolStudents ManagementCourses ExecutivePGinManagement MBAinDigitalMarketing&DataSciencefromJAIN OnlineMBAinShivNadar BachelorofBusinessfromDeakinUniversity BBADegreeFromJainUniversity ExecutiveMBAfromPES NUSFutureLeadersProgram PGinProductManagementandAnalytics BachelorofBusinessAnalyticsFromDeakin CloudComputingCourses PGPinCloudComputingCourse AdvancedSoftwareEngineeringCourseIITMadras CyberSecurityCourses PGPrograminCyberSecurity DigitalMarketing&SalesCourses PGPinStrategicDigitalMarketingCourse AdvancedCertificateinStrategicDigitalMarketing AdvancedCertificatePrograminSales SoftwareEngineeringCourses SoftwareDevelopmentIITRoorkee JainOnlineMCA AdvancedSoftwareEngineeringIITMadras DesignThinkingCourses StanfordExecutiveDesignThinking PGinUserExperienceDesignIIT-Bombay FreeCourses DataScienceFreeCourses DataScienceFoundations ProbabilityforDataScience LinearProgrammingforDataScience IntroductiontoR DataVisualizationusingTableau MachineLearningFreeCourses BasicsofMachineLearning StatisticsforMachineLearning FakeNewsDetectionusingMachineLearning ProductCategorizationusingMachineLearning PythonforMachineLearning ArtificialIntelligenceFreeCourses IntroductiontoArtificialIntelligence ArtificialIntelligenceProjects IntroductiontoDeepLearning FoundationsofAI IntroductiontoNaturalLanguageProcessing CyberSecurityFreeCourses IntroductiontoCyberSecurity TypesofCyberSecurity CyberSecurityThreats CyberSecurityForLeadership CyberForensics CloudComputingFreeCourses CloudFoundations CloudComputingServiceModels BestCloudCertifications CloudComputingArchitecture ApplicationsofCloudComputing BigDataFreeCourses IntroductiontoBigDataandHadoop MasteringBigDataAnalytics HadoopMapReduce IntroductiontoApacheHive SparkBasics MarketingFreeCourses IntroductiontoDigitalMarketing SEOforBeginners ContentMarketingBasics SearchEngineMarketing AffiliateMarketing EmailMarketing InterviewQuestions PythonInterviewQuestions JavaInterviewQuestions SQLInterviewQuestions MachineLearningInterviewQuestions SeleniumInterviewQuestions NLPInterviewQuestions C++InterviewQuestions AWSInterviewQuestions RPAInterviewQuestions CloudComputingInterviewQuestions Career CareerTopics CareerinAI Careerinpython CareerinCyberSecurity CareerinDigitalMarketer CareerinCloudArchitect CareerinBusinessAnalytics CareerinEthicalHacking Salary DataScientistSalary PythonDeveloperSalary CyberSecuritySalary DataAnalystSalary BusinessAnalystSalary MachineLearningSalary BlockchainDeveloperSalary DigitalMarketingSalary DevOpsSalary Skills DataScientistSkills BigDataAnalyticsSkills CloudArchitectSkills ManagementSkills CloudEngineerSkills AdministrativeSkills DigitalMarketingSkills DevOpsEngineerSkills Resume WebDeveloperResume BusinessAnalystResume DataScienceResume JavaDeveloperResume DigitalMarketingResume PythonDeveloperResume DevOpsEngineerResume ArtificialIntelligenceResume MachineLearningResume Jobs HighestPayingJobsintheWorld HighestPayingFinanceJobs HighestPayingFreelancingJobs HighestPayingITJobs Non–TechnicalJobs CoverLetter CareerBreakCoverLetter DevOpsCoverLetter ContentManagerCoverLetter SEOManagerCoverLetter AndroidDeveloperCoverLetter OtherResources ArtificialIntelligence MachineLearning PythonProgramming DataScience CloudComputing CyberSecurity BusinessManagement VisualStories Search AllCourses DataScienceCourses Master’sinDataScience M.TechinDataScience(Bangalore) DataScienceSpecialization PGinBusinessAnalytics DecisionMakingUsingDataScience DataScienceandMachineLearning ArtificialIntelligenceCourses PGP–ArtificialIntelligenceandMachineLearning(Online) PGPinArtificialIntelligenceandMachineLearning(Classroom) PGP–ArtificialIntelligenceforLeaders PGinMachineLearning PGDiplomainComputerScienceandAI–IIITDelhi AIforHighSchoolStudents ManagementCourses ExecutivePGinManagement MBAinDigitalMarketing&DataSciencefromJAIN OnlineMBAinShivNadar BachelorofBusinessfromDeakinUniversity BBADegreeFromJainUniversity ExecutiveMBAfromPES NUSFutureLeadersProgram PGinProductManagementandAnalytics BachelorofBusinessAnalyticsFromDeakin CloudComputingCourses PGPinCloudComputingCourse AdvancedSoftwareEngineeringCourseIITMadras CyberSecurityCourses PGPrograminCyberSecurity DigitalMarketing&SalesCourses PGPinStrategicDigitalMarketingCourse AdvancedCertificateinStrategicDigitalMarketing AdvancedCertificatePrograminSales SoftwareEngineeringCourses SoftwareDevelopmentIITRoorkee JainOnlineMCA AdvancedSoftwareEngineeringIITMadras DesignThinkingCourses StanfordExecutiveDesignThinking PGinUserExperienceDesignIIT-Bombay FreeCourses DataScienceFreeCourses DataScienceFoundations ProbabilityforDataScience LinearProgrammingforDataScience IntroductiontoR DataVisualizationusingTableau MachineLearningFreeCourses BasicsofMachineLearning StatisticsforMachineLearning FakeNewsDetectionusingMachineLearning ProductCategorizationusingMachineLearning PythonforMachineLearning ArtificialIntelligenceFreeCourses IntroductiontoArtificialIntelligence ArtificialIntelligenceProjects IntroductiontoDeepLearning FoundationsofAI IntroductiontoNaturalLanguageProcessing CyberSecurityFreeCourses IntroductiontoCyberSecurity TypesofCyberSecurity CyberSecurityThreats CyberSecurityForLeadership CyberForensics CloudComputingFreeCourses CloudFoundations CloudComputingServiceModels BestCloudCertifications CloudComputingArchitecture ApplicationsofCloudComputing BigDataFreeCourses IntroductiontoBigDataandHadoop MasteringBigDataAnalytics HadoopMapReduce IntroductiontoApacheHive SparkBasics MarketingFreeCourses IntroductiontoDigitalMarketing SEOforBeginners ContentMarketingBasics SearchEngineMarketing AffiliateMarketing EmailMarketing InterviewQuestions PythonInterviewQuestions JavaInterviewQuestions SQLInterviewQuestions MachineLearningInterviewQuestions SeleniumInterviewQuestions NLPInterviewQuestions C++InterviewQuestions AWSInterviewQuestions RPAInterviewQuestions CloudComputingInterviewQuestions Career CareerTopics CareerinAI Careerinpython CareerinCyberSecurity CareerinDigitalMarketer CareerinCloudArchitect CareerinBusinessAnalytics CareerinEthicalHacking Salary DataScientistSalary PythonDeveloperSalary CyberSecuritySalary DataAnalystSalary BusinessAnalystSalary MachineLearningSalary BlockchainDeveloperSalary DigitalMarketingSalary DevOpsSalary Skills DataScientistSkills BigDataAnalyticsSkills CloudArchitectSkills ManagementSkills CloudEngineerSkills AdministrativeSkills DigitalMarketingSkills DevOpsEngineerSkills Resume WebDeveloperResume BusinessAnalystResume DataScienceResume JavaDeveloperResume DigitalMarketingResume PythonDeveloperResume DevOpsEngineerResume ArtificialIntelligenceResume MachineLearningResume Jobs HighestPayingJobsintheWorld HighestPayingFinanceJobs HighestPayingFreelancingJobs HighestPayingITJobs Non–TechnicalJobs CoverLetter CareerBreakCoverLetter DevOpsCoverLetter ContentManagerCoverLetter SEOManagerCoverLetter AndroidDeveloperCoverLetter OtherResources ArtificialIntelligence MachineLearning PythonProgramming DataScience CloudComputing CyberSecurity BusinessManagement VisualStories More Search AllCourses DataScienceCourses Master’sinDataScience M.TechinDataScience(Bangalore) DataScienceSpecialization PGinBusinessAnalytics DecisionMakingUsingDataScience DataScienceandMachineLearning ArtificialIntelligenceCourses PGP–ArtificialIntelligenceandMachineLearning(Online) PGPinArtificialIntelligenceandMachineLearning(Classroom) PGP–ArtificialIntelligenceforLeaders PGinMachineLearning PGDiplomainComputerScienceandAI–IIITDelhi AIforHighSchoolStudents ManagementCourses ExecutivePGinManagement MBAinDigitalMarketing&DataSciencefromJAIN OnlineMBAinShivNadar BachelorofBusinessfromDeakinUniversity BBADegreeFromJainUniversity ExecutiveMBAfromPES NUSFutureLeadersProgram PGinProductManagementandAnalytics BachelorofBusinessAnalyticsFromDeakin CloudComputingCourses PGPinCloudComputingCourse AdvancedSoftwareEngineeringCourseIITMadras CyberSecurityCourses PGPrograminCyberSecurity DigitalMarketing&SalesCourses PGPinStrategicDigitalMarketingCourse AdvancedCertificateinStrategicDigitalMarketing AdvancedCertificatePrograminSales SoftwareEngineeringCourses SoftwareDevelopmentIITRoorkee JainOnlineMCA AdvancedSoftwareEngineeringIITMadras DesignThinkingCourses StanfordExecutiveDesignThinking PGinUserExperienceDesignIIT-Bombay FreeCourses DataScienceFreeCourses DataScienceFoundations ProbabilityforDataScience LinearProgrammingforDataScience IntroductiontoR DataVisualizationusingTableau MachineLearningFreeCourses BasicsofMachineLearning StatisticsforMachineLearning FakeNewsDetectionusingMachineLearning ProductCategorizationusingMachineLearning PythonforMachineLearning ArtificialIntelligenceFreeCourses IntroductiontoArtificialIntelligence ArtificialIntelligenceProjects IntroductiontoDeepLearning FoundationsofAI IntroductiontoNaturalLanguageProcessing CyberSecurityFreeCourses IntroductiontoCyberSecurity TypesofCyberSecurity CyberSecurityThreats CyberSecurityForLeadership CyberForensics CloudComputingFreeCourses CloudFoundations CloudComputingServiceModels BestCloudCertifications CloudComputingArchitecture ApplicationsofCloudComputing BigDataFreeCourses IntroductiontoBigDataandHadoop MasteringBigDataAnalytics HadoopMapReduce IntroductiontoApacheHive SparkBasics MarketingFreeCourses IntroductiontoDigitalMarketing SEOforBeginners ContentMarketingBasics SearchEngineMarketing AffiliateMarketing EmailMarketing InterviewQuestions PythonInterviewQuestions JavaInterviewQuestions SQLInterviewQuestions MachineLearningInterviewQuestions SeleniumInterviewQuestions NLPInterviewQuestions C++InterviewQuestions AWSInterviewQuestions RPAInterviewQuestions CloudComputingInterviewQuestions Career CareerTopics CareerinAI Careerinpython CareerinCyberSecurity CareerinDigitalMarketer CareerinCloudArchitect CareerinBusinessAnalytics CareerinEthicalHacking Salary DataScientistSalary PythonDeveloperSalary CyberSecuritySalary DataAnalystSalary BusinessAnalystSalary MachineLearningSalary BlockchainDeveloperSalary DigitalMarketingSalary DevOpsSalary Skills DataScientistSkills BigDataAnalyticsSkills CloudArchitectSkills ManagementSkills CloudEngineerSkills AdministrativeSkills DigitalMarketingSkills DevOpsEngineerSkills Resume WebDeveloperResume BusinessAnalystResume DataScienceResume JavaDeveloperResume DigitalMarketingResume PythonDeveloperResume DevOpsEngineerResume ArtificialIntelligenceResume MachineLearningResume Jobs HighestPayingJobsintheWorld HighestPayingFinanceJobs HighestPayingFreelancingJobs HighestPayingITJobs Non–TechnicalJobs CoverLetter CareerBreakCoverLetter DevOpsCoverLetter ContentManagerCoverLetter SEOManagerCoverLetter AndroidDeveloperCoverLetter OtherResources ArtificialIntelligence MachineLearning PythonProgramming DataScience CloudComputing CyberSecurity BusinessManagement VisualStories More Search HomeArtificialIntelligenceAIProjectsWhatisImageRecognitionandHowitisUsed? ArtificialIntelligenceAIProjectsDeepLearningNeuralNetworks Share FacebookTwitterWhatsApp WhatisImagerecognition?HowdoesImagerecognitionwork?WorkingofConvolutionalandPoolinglayersImagerecognitionusingPythonImagerecognitionwithapre-trainednetwork ThevisualperformanceofHumansismuchbetterthanthatofcomputers,probablybecauseofsuperiorhigh-levelimageunderstanding,contextualknowledge,andmassivelyparallelprocessing.Buthumancapabilitiesdeterioratedrasticallyafteranextendedperiodofsurveillance,alsocertainworkingenvironmentsareeitherinaccessibleortoohazardousforhumanbeings.Soforthesereasons,automaticrecognitionsystemsaredevelopedforvariousapplications.Drivenbyadvancesincomputingcapabilityandimageprocessingtechnology,computermimicryofhumanvisionhasrecentlygainedgroundinanumberofpracticalapplications. LearnwithIndia’s#1programinAIMLandbuildjob-readyskills.DesignedincollaborationwiththeUniversityofTexasatAustin,thisprogramoffersacomprehensivecurriculumtohelpprofessionalsupskillfast.Youwillpickupindustry-valuedskillsinalltheAIMLconceptslikeMachineLearning,ComputerVision,NaturalLanguageProcessing,NeuralNetworks,andmore.Thisprogramalsoincludesseveralguidedprojectstohelpyoubecomeexperts.GreatLearningalsoofferspersonalizedcareercoachingandinterviewpreparationtohelpyouacetherecruitingprocess. WhatisImagerecognition? Imagerecognitionreferstotechnologiesthatidentifyplaces,logos,people,objects,buildings,andseveralothervariablesindigitalimages.Itmaybeveryeasyforhumanslikeyouandmetorecognisedifferentimages,suchasimagesofanimals.Wecaneasilyrecognisetheimageofacatanddifferentiateitfromanimageofahorse.Butitmaynotbesosimpleforacomputer. Adigitalimageisanimagecomposedofpictureelements,alsoknownaspixels,eachwithfinite,discretequantitiesofnumericrepresentationforitsintensityorgreylevel.Sothecomputerseesanimageasnumericalvaluesofthesepixelsandinordertorecogniseacertainimage,ithastorecognisethepatternsandregularitiesinthisnumericaldata. Animageofadogrepresentedby40x40pixels. Imagerecognitionshouldnotbeconfusedwithobjectdetection.Inobjectdetection,weanalyseanimageandfinddifferentobjectsintheimagewhileimagerecognitiondealswithrecognisingtheimagesandclassifyingthemintovariouscategories. HowdoesImagerecognitionwork? Typicallythetaskofimagerecognitioninvolvesthecreationofaneuralnetworkthatprocessestheindividualpixelsofanimage.Thesenetworksarefedwithasmanypre-labelledimagesaswecan,inorderto“teach”themhowtorecognizesimilarimages. Soletmebreaktheprocessforyouinsomesimplesteps: Weneedadatasetcontainingimageswiththeirrespectivelabels.Forexample,animageofadogmustbelabelledasadogorsomethingthatwecanunderstand.Next,theseimagesaretobefedintoaNeuralNetworkandthentrainedonthem.Usually,forthetasksconcernedwithimages,weuseconvolutionalneuralnetworks.ThesenetworksconsistofconvolutionallayersandpoolinglayersinadditiontoMultiperceptronlayers(MLP).Theworkingofconvolutionalandpoolinglayersareexplainedinthebelow.Wefeedintheimagethatisnotinthetrainingsetandgetpredictions. Inthecomingsections,byfollowingthesesimplestepswewillmakeaclassifierthatcanrecogniseRGBimagesof10differentkindsofanimals. Note:Themodelwillonlybeabletorecogniseanimalsthatareinthedataset.Forexample,amodeltrainedtorecognisedogsandcatcannotrecogniseboats  WorkingofConvolutionalandPoolinglayers ConvolutionallayersandPoolinglayersarethemajorbuildingblocksusedinconvolutionalneuralnetworks.Letusseethemindetail HowdoesConvolutionalLayerwork? Theconvolutionallayer’sparametersconsistofasetoflearnablefilters(orkernels),whichhaveasmallreceptivefield.Thesefiltersscanthroughimagepixelsandgatherinformationinthebatchofpictures/photos.Convolutionallayersconvolvetheinputandpassitsresulttothenextlayer.Thisisliketheresponseofaneuroninthevisualcortextoaspecificstimulus.  Convolutionoperation Belowisanexampleofhowconvolutionoperationisdoneonanimage.Asimilarprocessisdoneforallthepixels. Hereisanexampleofanimageinourtestsetthathasbeenconvolutedwithfourdifferentfiltersandhencewegetfourdifferentimages. HowdoesPoolingLayerwork? Thepoolingoperationinvolvesslidingatwo-dimensionalfilterovereachchannelofthefeaturemapandsummarisingthefeatureslyingwithintheregioncoveredbythefilter.Apoolinglayerisusuallyincorporatedbetweentwosuccessiveconvolutionallayers.Thepoolinglayerreducesthenumberofparametersandcomputationbydown-samplingtherepresentation.Thepoolingfunctioncanbeeithermaxoraverage.Maxpoolingiscommonlyusedasitworksbetter Thepoolingoperationinvolvesslidingatwo-dimensionalfilterovereachchannelofthefeaturemapandsummarisingthefeatureslyingwithintheregioncoveredbythefilter.Thisprocessisillustratedbelow. Whenpassingthefourimageswegotafterconvolutionthroughamax-poolinglayerofdimension2×2,wegetthisasoutput Aswecansee,thedimensionshavedecreasedbyonehalfbuttheinformationintheimageisstillpreserved. ImagerecognitionusingPython HereIamgoingtousedeeplearning,morespecificallyconvolutionalneuralnetworksthatcanrecogniseRGBimagesoftendifferentkindsofanimals.AnRGBimagecanbeviewedasthreedifferentimages(aredscaleimage,agreenscaleimageandabluescaleimage)stackedontopofeachother,andwhenfedintothered,greenandblueinputsofacolourmonitor,itproducesacolourimageonthescreen.Weuseadatasetknownas Animals-10fromKaggle. So,letusstartmakingaclassifierusingPythonandKeras.WearegoingtoimplementtheprograminColabasweneedalotofprocessingpowerandGoggleColabprovidesfreeGPUs.Theoverallstructureoftheneuralnetworkwearegoingtousecanbeseeninthisimage TheveryfirststepistogetdataonyourColabnotebook.Youdon’tneedahigh-speedinternetforthisasitisdirectlydownloadedintogooglecloudfromtheKagglecloud. Forgettingthedata,followthesesteps: GotoyourKaggleaccountandclickonmyaccounts.Incaseyoudon’thaveaKaggleaccount,createone,itisfree.Next,downloadthekaggle.jsonfilebyclickingonthebutton‘createnewAPItoken’.GotoyourColabnotebookandstartcoding Inthistutorial,weareusingImageGeneratortolabeltheimages.So,incaseyouareusingsomeotherdataset,besuretoputallimagesofthesameclassinthesamefolder.Andthenplaceallthefoldersinthefolder. #Thesestepsaretobefollowedwhenusinggooglecolab #andimportingdatafromkaggle fromgoogle.colabimportfiles #InstallKagglelibrary !pipinstall-qkaggle fromgoogle.colabimportfiles #uploadthekaggle.jsonfile uploaded=files.upload() #makeadiectoryinwhichkajggle.jsonisstored #!mkdir~/.kaggle !cpkaggle.json~/.kaggle/ #downloadthedatasetintothecolab !kaggledatasetsdownload-dalessiocorrado99/animals10 #unzipthedata !unzip/content/animals10.zip #Incaseyouareusingalocalmachine,startfromhere. importtensorflowastf fromtensorflow.keras.preprocessing.imageimportImageDataGenerator fromtensorflow.keras.layersimportInput,Dense fromtensorflow.kerasimportSequential,Model fromtensorflow.keras.layersimportBatchNormalization,Dropout,Flatten fromtensorflow.keras.layersimportConv2D fromtensorflow.keras.layersimportMaxPooling2D fromtensorflow.keras.layersimportGlobalAveragePooling2D fromtensorflow.keras.preprocessingimportimage importnumpyasnp importos importcv2 train_data_dir='/kaggle/input/animals10/raw-img/' img_height=128 img_width=128 batch_size=64 nb_epochs=20 train_datagen=ImageDataGenerator(rescale=1./255, shear_range=0.2, zoom_range=0.2, horizontal_flip=True, validation_split=0.2)#setvalidationsplit train_generator=train_datagen.flow_from_directory( train_data_dir, target_size=(img_height,img_width), batch_size=batch_size, class_mode='categorical', subset='training')#setastrainingdata validation_generator=train_datagen.flow_from_directory( train_data_dir,#samedirectoryastrainingdata target_size=(img_height,img_width), batch_size=batch_size, class_mode='categorical', subset='validation')#setasvalidationdata model=Sequential() inputShape=(128,128,3) model.add(Conv2D(64,(3,3),padding="same",activation='relu',input_shape=inputShape)) model.add(BatchNormalization()) model.add(Conv2D(32,kernel_size=5,strides=2,padding='same',activation='relu')) model.add(MaxPooling2D((2,2))) model.add(Dropout(0.4)) model.add(Conv2D(64,kernel_size=5,strides=2,padding='same',activation='relu')) model.add(MaxPooling2D((2,2))) model.add(BatchNormalization()) model.add(Dropout(0.4)) model.add(Flatten()) model.add(Dropout(0.4)) model.add(Dense(64,activation='relu')) model.add(BatchNormalization()) model.add(Dense(10,activation='softmax')) model.summary() #compilethemodel model.compile(optimizer="adam",loss="categorical_crossentropy",metrics=["accuracy"]) #trainthemodel,thissteptakesalotoftime(hours) model.fit_generator( train_generator, steps_per_epoch=train_generator.samples//batch_size, validation_data=validation_generator, validation_steps=validation_generator.samples//batch_size, epochs=nb_epochs) #savethemodelforlateruse model.save('path\nameofmodel') #orderoftheanimalsarrayisimportant #animals=["dog","horse","elephant","butterfly","chicken","cat","cow","sheep","spider","squirrel"] bio_animals=sorted(os.listdir('/content/raw-img')) categories={'cane':'dog',"cavallo":"horse","elefante":"elephant","farfalla":"butterfly","gallina":"chicken","gatto":"cat","mucca":"cow","pecora":"sheep","scoiattolo":"squirrel","ragno":"spider"} defrecognise(pred): animals=[categories.get(item,item)foriteminbio_animals] print("Theimageconsistof",animals[pred]) fromtensorflow.keras.preprocessingimportimage importnumpyasnp img=image.load_img("https://d1m75rqqgidzqn.cloudfront.net/kaggle/input/testttt/OIF-e2bexWrojgtQnAPPcUfOWQ.jpeg",target_size=(128,128)) x=image.img_to_array(img) x=np.expand_dims(x,axis=0) prediction=model.predict(x) #prediction recognise(np.argmax(prediction)) test_data_path="/content/testdata/test_animals" files=sorted(os.listdir(test_data_path)) files=files[1:] forimginfiles: x=cv2.imread(os.path.join(test_data_path,img)) cv2_imshow(x) recognise(np.argmax(predict[files.index(img)])) print("") Output:Idownloadedsomeimagesfromgoogleandusedthismodeltolabelthem.Herearetheresults TopredictImages,weneedtouploadthemtotheColab(getsdeletedautomaticallyafterthesessionisended)oryoucanevendownloadthemtoyourgoogledrivepermanently. Followthestepsbelowtocreateadirectoryfortestdata CreateanewfoldercalledtestdataNext,createanotherfolderinthisfoldernamedtestanimalsUploadyourimagestothisfolder. Aswecansee,thismodeldidadecentjobandpredictedallimagescorrectlyexcepttheonewithahorse.Thisisbecausethesizeofimagesisquitebigandtogetdecentresults,themodelhastobetrainedforatleast100epochs.Butduetothelargesizeofthedatasetandimages,Icouldonlytrainitfor20epochs(took4hoursonColab). Toincreasetheaccuracyandgetanaccurateprediction,wecanuseapre-trainedmodelandthencustomisethataccordingtoourproblem. ImageRecognitionwithapre-trainedmodel Inthisexample,IamgoingtousetheXceptionmodelthathasbeenpre-trainedonImagenetdataset.ThistechniqueisbasicallycalledTransferlearning. XceptionModelisproposedbyFrancoisChollet.XceptionisanextensionoftheinceptionArchitecturewhichreplacesthestandardInceptionmoduleswithdepthwiseSeparableConvolutions.ThismodelisavailableonKerasandwejustneedtoimportit.Solet’sstartcoding fromgoogle.colabimportfiles #InstallKagglelibrary !pipinstall-qkaggle fromgoogle.colabimportfiles #uploadthekaggle.jsonfile uploaded=files.upload() #makeadiectoryinwhichkajggle.jsonisstored #!mkdir~/.kaggle !cpkaggle.json~/.kaggle/ #downloadthedatasetintothecolab !kaggledatasetsdownload-dalessiocorrado99/animals10 #unzipthedata !unzip/content/animals10.zip importtensorflowastf fromtensorflow.keras.preprocessing.imageimportImageDataGenerator fromtensorflow.keras.layersimportInput,Dense fromtensorflow.kerasimportSequential,Model fromtensorflow.keras.layersimportBatchNormalization,Dropout,Flatten fromtensorflow.keras.layersimportConv2D fromtensorflow.keras.layersimportMaxPooling2D fromtensorflow.keras.layersimportGlobalAveragePooling2D fromtensorflow.keras.preprocessingimportimage fromtensorflow.keras.layersimportGlobalAveragePooling2D importnumpyasnp importos importcv2 train_data_dir='/kaggle/input/animals10/raw-img/' img_height=299 img_width=299 batch_size=64 nb_epochs=20 train_datagen=ImageDataGenerator(rescale=1./255, shear_range=0.2, zoom_range=0.2, horizontal_flip=True, validation_split=0.2)#setvalidationsplit train_generator=train_datagen.flow_from_directory( train_data_dir, target_size=(img_height,img_width), batch_size=batch_size, class_mode='categorical', subset='training')#setastrainingdata validation_generator=train_datagen.flow_from_directory( train_data_dir,#samedirectoryastrainingdata target_size=(img_height,img_width), batch_size=batch_size, class_mode='categorical', subset='validation')#setasvalidationdata #importapre-trainedmodel,withoutthetoplayers.Wewillcustomise #thetoplayersforourproblem base_model=tf.keras.applications.Xception(include_top=False,input_shape=(299,299,3)) #Fornowfreezetheinitiallayersanddonottrainthem forlayerinbase_model.layers: layer.trainable=False #createacustomtopclassifier x=base_model.output x=GlobalAveragePooling2D()(x) x=Dense(516,activation='relu')(x) #sinceourproblemhas10differntanimalswehave10classes #thuswekeep10nodesinthelastlayer predictions=Dense(10,activation='softmax')(x) model=Model(inputs=base_model.inputs,outputs=predictions) model.summary() model.compile(optimizer="adam",loss="categorical_crossentropy",metrics=["accuracy"]) model.fit_generator( train_generator, steps_per_epoch=train_generator.samples//batch_size, validation_data=validation_generator, validation_steps=validation_generator.samples//batch_size, epochs=nb_epochs) #Nowunfreezethelayersandtrainthewholemodel forlayerinbase_model.layers: layer.trainable=True history=model.fit_generator( train_generator, steps_per_epoch=train_generator.samples//batch_size, validation_data=validation_generator, validation_steps=validation_generator.samples//batch_size, epochs=nb_epochs) model.save('path\nameofmodel') #orderoftheanimalsarrayisimportant #animals=["dog","horse","elephant","butterfly","chicken","cat","cow","sheep","spider","squirrel"] bio_animals=sorted(os.listdir('/content/raw-img')) categories={'cane':'dog',"cavallo":"horse","elefante":"elephant","farfalla":"butterfly","gallina":"chicken","gatto":"cat","mucca":"cow","pecora":"sheep","scoiattolo":"squirrel","ragno":"spider"} defrecognise(pred): animals=[categories.get(item,item)foriteminbio_animals] print("Theimageconsistof",animals[pred]) fromtensorflow.keras.preprocessingimportimage importnumpyasnp img=image.load_img("https://d1m75rqqgidzqn.cloudfront.net/kaggle/input/testttt/OIF-e2bexWrojgtQnAPPcUfOWQ.jpeg",target_size=(299,299)) x=image.img_to_array(img) x=np.expand_dims(x,axis=0) prediction=model.predict(x) #prediction recognise(np.argmax(prediction)) test_data_path="/content/testdata/test_animals" files=sorted(os.listdir(test_data_path)) files=files[1:] forimginfiles: x=cv2.imread(os.path.join(test_data_path,img)) cv2_imshow(x) recognise(np.argmax(predict[files.index(img)])) print("") Output: Aswecanseethemodelmakesaccuratepredictionsonallofthedatainourtestdataset.Ihavesavedthismodel,henceitcanbeusedatanytimebyusingthefunctionshownbelow: fromtensorflowimportkeras model=keras.models.load_model('path.h5') #e.g.model=keras.models.load_model('/content/simpleconvkag.h5') Incaseyouwantthecopyofthetrainedmodelorhaveanyqueriesregardingthecode,feelfreetodropacomment. Thisbringsustotheendofthisarticle.Wehavelearnedhowimagerecognitionworksandclassifieddifferentimagesofanimals. IfyouwishtolearnmoreaboutPythonandtheconceptsofMachinelearning,upskillwithGreatLearning’sPGProgramArtificialIntelligenceandMachineLearning. RELATEDARTICLESMOREFROMAUTHOR ArtificialIntelligence TopDeepLearningCoursesOnline-2022 ComputerVision Real-timeFacedetection|FaceMaskDetectionusingOpenCV ComputerVision HowToApplyMachineLearningtoRecogniseHandwriting|HowtoRecognisehandwriting 2COMMENTS heyhowdidyoudecidearrayorderingoflabels Reply HiAshutosh, ItdependsonthedatasetweareworkingonandheremydatasetcontainedfoldersofeachanimalintheorderIhavementionedhere.Ifyouarefacingproblemsregardingthelabels,itismostprobablythatyourfoldersarenotinthesameorderasitwasinmycase.Youcanchangethelabelarrayaccordingtotheorderofyourfolders. Reply LEAVEAREPLYCancelreply Pleaseenteryourcomment! Pleaseenteryournamehere Youhaveenteredanincorrectemailaddress! Pleaseenteryouremailaddresshere Savemyname,email,andwebsiteinthisbrowserforthenexttimeIcomment. Pleaseenterananswerindigits:two×4= Δ GreatLearningisaned-techcompanythatoffersimpactfulandindustry-relevantprogramsinhigh-growthareas.Withastrongpresenceacrosstheglobe,wehaveempowered10,000+learnersfromover50countriesinachievingpositiveoutcomesfortheircareers.KnowMore FacebookInstagramLinkedinPinterestTwitter ©2021GreatLearningAllrightsreserved