Improved Automated Transcripts

Rajiv PoddarLeave a comment

Our latest speech and language models have been released. There are several new features in this release. The following is a list:

Acoustic Model: This is our fourth acoustic model trained on our data. The dataset contained mostly accented speakers (eg. Indian, African, Irish etc.). It also contained some noisy files. The accuracy of the automated transcript on accented files should be better now.

Language Model: We have added more data to our language model and doubled its size. The model now model has now been trained on around 46 million lines and has improved the WER by around 2%.

Punctuations: The biggest feature of this release is expanded punctuations. We now support all types of punctuations including quotes and hyphens. To our knowledge, nobody else including Google Web Speech, AWS Transcribe and Speechmatics supports quotes.

Speaker Turns: We also have updated our speaker turns model. The accuracy of the model is around 80% on long paragraphs. The automated transcripts will be better segmented now. We are currently working on adding speaker diarization to the automated transcript and it should be out soon. We do speaker turns a bit differently and do not require the number of speakers as an input. That is also one of our unique features. Google Web Speech does not support multi-speaker files and AWS Transcribe and Speechmatics require the number of speakers as an input for diarization.

This release also fixes the issue of missing predictions where some words, especially near speaker turns were not being transcribed. The automated transcripts should now capture all utterances, except filler words. We also benchmarked our model with LibriSpeech Clean and our internal dataset. The following are our numbers.

Dataset Type WER CER
LibriSpeech Clean Read speech 14.74% 5.96%
Scribie Internal Conversational 16.33% 8.82%

For comparison, PaddlePaddle numbers are the following:

Dataset Type WER CER
LibriSpeech Clean Read speech 5.4% 1.9%
Scribie Internal Conversational 22.68% 12.32%

As you can see, for conversational audio, our models outperform PaddlePaddle by a wide margin. We are working on improving our models for non-conversational audio as well. Our ASR is a DeepSpeech-based system and therefore a comparison with PaddlePaddle is a good benchmark for us. The Continual Learning blog post has some more details on how we trained our DeepSpeech models.

The automated transcripts are free currently, so try it out today!

Continual Learning for Speech-to-Text

Rajiv PoddarLeave a comment

Flawless transcripts and fast turnaround time are the hallmarks of Scribie. Not only are our transcripts highly accurate, but also priced reasonably. But have you ever wondered what makes that possible? The answer lies in constantly improving our speech-to-text engine, which assists our transcribers. We provide automatic word completion to our transcribers, and the better those autocompletes are, the less they have to type.

Our speech recognition engine is a Deep Learning system. For the uninitiated, Deep Learning is a subdomain of Machine Learning. It makes use of Artificial Neural Networks that, in a way, mimic the structure and function of the human brain. Our speech recognition engine is based on the DeepSpeech 2 network from Baidu, and written in PyTorch.

Scribie has a large dataset of audio and transcripts — over 100,000 hours at the last count. Training Deep Learning models over such a large dataset is very expensive in practice, as it requires a large number of GPUs and SSDs. For comparison, Baidu trained their models with 256 GPUs on custom hardware when they developed the DeepSpeech architecture. We don’t have the time or money to do that. So we developed an approach which we call Continual Learning.

Continual Learning

We first built and trained a large model with a 3,000-hour dataset. That took around three weeks on our rig. Since then, every month we have built a ‘corrections dataset’ of around 1,000 hours. This corrections dataset is made up of predictions from the previous model that were wrong, and then manually corrected by our transcribers. In each iteration we remove an equal amount of data from the previous training set and fine-tune the model over the newly combined data. This ensures that our model keeps improving over time.

Results

We have completed three such iterations and the results are promising. We have been able to consistently decrease the Word Error Rate, a common metric for automated transcription accuracy. The following is the chart of our WER.

We are providing free automated transcripts for a limited time, so please don’t hesitate to try out our online speech recognition system soon!

Deep Learning and AI has been in the news a lot lately, and there are concerns that Machine Learning will end up taking our jobs and replace humans. We have taken a different approach and built a system to assist our transcribers instead. Eventually, we want to reach a point where a human would have to spend just 10 minutes on a one-hour file, and still produce a highly accurate transcript of it. We still have a long way to go and we are working hard at it!

Did Trump Say I or I’d?

Rajiv PoddarLeave a comment

The Jury is out on what President Trump said: I or I’d. But what does the AI think? We put our free automated transcripts service to test on the following clip.

And here’s the result.

But with that being said, a president has been extremely generous with what he said. I like him a lot. I have a great relationship with them, as you know, have a great relationship with prime minister abe in japan, and I probably have a very good relationship with m gun f not care. I have relationships with people to surprise.

So our AI agrees with WSJ. President Trump did say ‘I’. So there you go!

The transcript is missing few words towards the end and we are working to fix it. However, if you have a clean audio file then head here to get a free automated transcript!

Introducing Billing

Rajiv PoddarLeave a comment

We are happy to announce that we now support billing accounts on Scribie.com. Billing accounts are where you can order your transcripts online and pay on a Net 15 or Net 30 basis. We send you a bill at the end of the month or whenever you request it. The volume requirement for billing accounts are higher though. We can only consider it for order amounts of more than $1000. We also require a contract to be signed before the billing account can be set up.

We were unable to support billing accounts previously as we followed a different model where we paid our transcribers as soon as their work was reviewed. That meant that we had to charge our customers upfront. This was radically different from other freelance marketplaces where there are restrictions on withdrawal of earnings. We decided to do away with such restrictions as our aim was to build the best place for audio/video transcription.

But Net billing is an important requirement for Enterprises and SMEs and many of our customers requested it. Our solution to this problem was to get a line of credit from our bank. We finally got the approval for it last week. A big shout out to our bankers for this!

So if you are looking for a billing account with us, just get in touch with us and we will start the process.

How to get your free automated transcripts from Scribie

Rajiv PoddarLeave a comment
Free Transcription Service

Getting your automated transcripts from Scribie is a fairly intuitive process. But here is a step-by-step guide on it.

  • Go to scribie.com/transcription/free
  • Click on “Upload Files” button.
  • You have many ways to add your file. Once you have chosen wait for the file to be uploaded.
  • Once you are on this page, click on “Auto”

    Upload Page

  • You will get a pop up to create an account or to sign-in if you already have an account.
  • Once you click on “Auto” again, you will get this message.

  • Wait for the email from Scribie on yourom registered mail id on the status of your file.
  • Once you get the confirmation mail, click on the link provided and you will be directed to the screen with the button ” Edit transcript”.

  • Click on “Edit Transcript” to start editing.

 

 

Price Cut

Rajiv Poddar1 Comment

We are pleased to announce an across-the-board 20% drop in our pricing effective today. Our new transcription rates are as follows:

Scribie New Transcription Rates
Old New Savings
Budget $0.75/min $0.60/min 20%
Regular $1.50/min $1.20/min 20%
Express $3.00/min $2.40/min 20%

We started with the mission to build the best place for transcription; both for transcribers and customers. We have been relentlessly pursuing our goal and recently have built technology that helps reduce the time and effort of transcribers, without compromising accuracy in any way. We are happy to pass on the savings to our customers.

We have always stood for accuracy and our goal has been to provide the highest quality transcript, at the lowest possible cost. However, we still want to compensate our transcribers fairly. The only way to solve this problem was with technology. Our tech has now been rolled out in production and we are happy to reach this milestone. This is the real test of whether our tech is good enough or not!

We will be talking more about our tech here in the coming days. So check back here if you’re interested in the details. In the meantime, upload your files online and order transcripts online to enjoy the benefits tech can offer with our reduced pricing.

 

No Room for Errors

Rajiv PoddarLeave a comment
No room for errors

If you have used Scribie’s service before, you probably know about our high standards in terms of the quality and the accuracy of the transcripts.

So a very important question that comes to mind is, how is Scribie able to churn out 99.9% accurate transcripts all the time, while some industry players are afraid of even claiming that benchmark? Continue reading “No Room for Errors” →

Simple Ways to Promote Your Podcast

Rajiv PoddarLeave a comment
Simple Ways to Promote Your Podcast

Do you have a podcast that you are putting a lot of effort into, but are not seeing any returns?

Do you think lack of traffic to your podcast is not motivating you to put the effort into producing useful content?

If your answer is yes to both those questions then you might want to know that you are not alone.

There are a lot of podcasters and other content creators who have been in the same situation that you are in right now but have managed to come out of that phase.

And in this post, we will look at the 5 factors that are common among a lot of these successful content creators. Continue reading “Simple Ways to Promote Your Podcast” →

Are Automated Transcribing Softwares Good Enough? Not for New York Times

Rajiv PoddarLeave a comment

If you are ChristineMcM, a New York Times commentator you probably know a too much about how automatic transcribing software can mess things up for you.

As reported by The Daily Dot, she had something to say about a recent Trump article but had to take a phone call in the middle of her comment. Her automatic transcription software heard and posted the whole conversation.

Yes, you read that right.

This is what it ended up posting.

This might be funny, but this shows the state that we are currently in with respect to automated transcribing.

Transcribing still continues to be mostly done by humans to avoid such gaffes.

Although she later clarified the mistake, it left those close to her and her followers in a state of a fix. Some even suspected that she might be having a neurological episode.

Here is her clarification:

Having understood these problems, Scribie is not looking to go the same route.

Instead, we use technology and AI to help humans transcribe faster and better.

The industry is far away from completely eliminating the human factor in the transcribing chain (unless you can afford such a gaffe).

For the time being the best way to get your file transcribed is a human with cutting-edge technology that enables efficiency and high accuracy.