Closed ovninosa closed 5 years ago
HiDid you look at the scores ?The default decision threshold is 0 I think.Not good (at least without score norm). Easy to fixJfEnvoyé depuis mon smartphone Samsung Galaxy. -------- Message d'origine --------De : ovninosa notifications@github.com Date : 15/11/2018 21:12 (GMT+01:00) À : ALIZE-Speaker-Recognition/android-alize-demo android-alize-demo@noreply.github.com Cc : Subscribed subscribed@noreply.github.com Objet : [ALIZE-Speaker-Recognition/android-alize-demo] Audio length (#2) Hello, I recently use the app, run OK. I can train my voice speaking between 5, 10 and 20 seconds. The response is always negative over test (verification) and identification. There is a good length for the training, verification and identification? Something like: Training: at least 20 seconds Verification: at least 5 seconds Identification: at least 10 seconds. Thanks
—You are receiving this because you are subscribed to this thread.Reply to this email directly, view it on GitHub, or mute the thread. {"api_version":"1.0","publisher":{"api_key":"05dde50f1d1a384dd78767c55493e4bb","name":"GitHub"},"entity":{"external_key":"github/ALIZE-Speaker-Recognition/android-alize-demo","title":"ALIZE-Speaker-Recognition/android-alize-demo","subtitle":"GitHub repository","main_image_url":"https://assets-cdn.github.com/images/email/message_cards/header.png","avatar_image_url":"https://assets-cdn.github.com/images/email/message_cards/avatar.png","action":{"name":"Open in GitHub","url":"https://github.com/ALIZE-Speaker-Recognition/android-alize-demo"}},"updates":{"snippets":[{"icon":"DESCRIPTION","message":"Audio length (#2)"}],"action":{"name":"View Issue","url":"https://github.com/ALIZE-Speaker-Recognition/android-alize-demo/issues/2"}}} [ { "@context": "http://schema.org", "@type": "EmailMessage", "potentialAction": { "@type": "ViewAction", "target": "https://github.com/ALIZE-Speaker-Recognition/android-alize-demo/issues/2", "url": "https://github.com/ALIZE-Speaker-Recognition/android-alize-demo/issues/2", "name": "View Issue" }, "description": "View this Issue on GitHub", "publisher": { "@type": "Organization", "name": "GitHub", "url": "https://github.com" } }, { "@type": "MessageCard", "@context": "http://schema.org/extensions", "hideOriginalBody": "false", "originator": "AF6C5A86-E920-430C-9C59-A73278B5EFEB", "title": "Audio length (#2)", "sections": [ { "text": "", "activityTitle": "ovninosa", "activityImage": "https://assets-cdn.github.com/images/email/message_cards/avatar.png", "activitySubtitle": "@ovninosa", "facts": [ { "name": "Repository: ", "value": "ALIZE-Speaker-Recognition/android-alize-demo" }, { "name": "Issue #: ", "value": 2 } ] } ], "potentialAction": [ { "name": "Add a comment", "@type": "ActionCard", "inputs": [ { "isMultiLine": true, "@type": "TextInput", "id": "IssueComment", "isRequired": false } ], "actions": [ { "name": "Comment", "@type": "HttpPOST", "target": "https://api.github.com", "body": "{\n\"commandName\": \"IssueComment\",\n\"repositoryFullName\": \"ALIZE-Speaker-Recognition/android-alize-demo\",\n\"issueId\": 2,\n\"IssueComment\": \"{{IssueComment.value}}\"\n}" } ] }, { "name": "Close issue", "@type": "HttpPOST", "target": "https://api.github.com", "body": "{\n\"commandName\": \"IssueClose\",\n\"repositoryFullName\": \"ALIZE-Speaker-Recognition/android-alize-demo\",\n\"issueId\": 2\n}" }, { "targets": [ { "os": "default", "uri": "https://github.com/ALIZE-Speaker-Recognition/android-alize-demo/issues/2" } ], "@type": "OpenUri", "name": "View on GitHub" }, { "name": "Unsubscribe", "@type": "HttpPOST", "target": "https://api.github.com", "body": "{\n\"commandName\": \"MuteNotification\",\n\"threadId\": 411654148\n}" } ], "themeColor": "26292E" } ]
Hi ,
I manage to get positive scoring with speaker model trained with my own voice (10 seconds) against verification (5 seconds). Issue I facing is the app will crash usually after N time of audio recording activity, and all the trained speaker model is gone after the app crash and not able to get it back when I rerun the app.
Regards, YiYang
Over my test I tried with audios from 5 seconds to 30 seconds on enrolment then 5 seconds and 10 seconds to test, always the threshold was negative. I did it wrong? or what I miss something?
THanks
Hello everyone, First of all, you can now set the recognition threshold as you like in the settings of the application. Secondly, I advise you to have a longer record to improve the results. Also, if you still are experiencing problems with the loss of the speakers models, you can save them in the settings.
Best regards, Nicolas.
Hello,
I recently use the app, run OK. I can train my voice speaking between 5, 10 and 20 seconds. The response is always negative over test (verification) and identification.
There is a good length for the training, verification and identification? Something like:
Training: at least 20 seconds Verification: at least 5 seconds Identification: at least 10 seconds.
Thanks