Closed

ASR / speech recognition expert for a custom ASR service

I need an expert in ASR / speech recognition to create a custom ASR service for us based on Mozilla DeepSpeech (or similiar open source technology - this is up to you).

The language recognition should be in GERMAN. ONLY GERMAN!

You should use all the online available training models to get the best results. Additionaly I can offer a database with 25 hours of high quality speech files + text for training as well.

Please only apply if you have already built and operated such systems yourself. In the end you should have an application that runs on my own servers, does not depend on external services and provides reliable and fast speech recognition.

But there is one special feature: For each audio file sent, the expected text is available. So there are two input parameters: The audio file and a string with the expected content.

The program only has to reliably determine whether the words in the string also occur in the audio file (in the correct order, of course, preferably with a score). The best result would be the following:

word1 : 1.0 (best recognition result)

word2 : 1.0 (best recognition result)

word3 : 0.5 (not sure result)

word4: 0 (missing)

So it is not a direct speech recognition, but rather a check, because the expected result is always known.

I offer a VM (ubuntu) in a data center to run the software. I would prefer a REST-API (http-post) for the requests.

Skills: Programming, Audio Processing, Linux, Software Development

See more: transcription online, automatic transcription software, transcribeme machine express, machine transcription service, speech to text software, automatic speech recognition, speech to text api, audio to text

About the Employer:
( 4 reviews ) Bielefeld, Germany

Project ID: #24607988

8 freelancers are bidding on average €1432 for this job

A2Design

Hello! We are A2 Design, a Russian-Canadian web development company. Having the required skills, we will be glad to help you with your project. We have some questions for you to clear up before we start. Please m More

€1500 EUR in 15 days
(6 Reviews)
6.3
akkastech

Hello, I hope you’re doing well. Thanks for the job post, it’s a pleasure meeting you. We’ve gone through your description, it's described pretty well about what you need. I am glad to share that we have experience in More

€2500 EUR in 25 days
(9 Reviews)
5.2
gkbhardwaj87

Hi, Nice to meet you. I have 7 year expereince in Machine learning and AI. I developed many application in AI using AWS ML services. For more information ping me.

€1125 EUR in 8 days
(1 Review)
3.1
intelliresponse

Dear Sir. My 10+ years of R&D experience should be of interest to an ideal project like yours. In various capacities in speech/image processing, leading to my present assignment as project manager. I have acquired skil More

€1000 EUR in 15 days
(1 Review)
2.6
gargankit642

Nice to meet you I am an Amazon Cloud Architect for the web infrastructure serving 90 million page impressions and 12 TB Internet traffic per month. The AWS services I use are EC2, ELB, MySQL RDS, VPC, CloudFront, Elas More

€1405 EUR in 20 days
(0 Reviews)
0.0
Alfatahbima9

********PLEASE GIVE ME CHANCE TO SHOW MY EXTRA ABILITY************ Thanks for your posting! I am a speech recognition and image processing expert using machine learning, such as Kaldi, tensorflow, caffe, darknet and e More

€1125 EUR in 7 days
(0 Reviews)
0.0
freelancermaddi9

Hello,I am Mashud,expert in deep learning, speech recognition and python. I have read the project description fully. I can complete your task within 7 days perfectly. I can assure you that you will get your job done e More

€1300 EUR in 7 days
(1 Review)
0.0
andrewlevin44

Hey, Interesting Job. It is not just simple STT, nor KWS. Rather similar to forced alignment. I'd love to work on it with Kaldi. Cheers

€1500 EUR in 14 days
(0 Reviews)
0.0