NeuralSpace Release Notes: What’s new in Version 1.5.0?
We are constantly developing the NeuralSpace Platform to give our users the best language AI capabilities. We’ve packaged these together in our latest release, which we have called Tony Tony Chopper.
In this version release, we bring to you a new service and some major Platform upgrades. Scroll down to catch up!
New AutoNLP Pipeline for NLU
Our AutoNLP pipeline has been updated which is faster and more accurate. This new pipeline boasts improved accuracy evaluated over 5 datasets. We also introduce a new fast mode for lightning-fast training at no additional cost.
In the next release, you will also have the option to choose the domain of the dataset (for example medical, finance, etc.) along with your desired language to get better accuracies in that domain.
Annotated Translation
When you translate a sentence with tagged entities, you lose the tags in the translated text as the words get rearranged. It is hard to find which phrase in the translated text corresponds to which entity in the original text.
With this service, you can pass entity annotations to our translation API and preserve these entities in the translated text.
With this release, we bring to you an improved version of the Annotated Translation service where you can tag multiple entities and retain them in the translated text with high accuracy. Check out a quick demo here.
Text-to-Speech (TTS)
We have added hyper-realistic voices to our already large TTS voice collection along with Natural Voice Generation capabilities.
Our TTS services will also be available on the no-code NeuralSpace Platform in the upcoming release.
Here is a sample audio of David Attenborough speaking about NeuralSpace’s mission statement.
*Note: these voices are not meant for any commercial use and are cloned and used for demo purposes only.
Vocab Adaptation
We introduce a new vocab adaption service that can improve STT model performance for any domain (especially proper nouns) using just text data. These improved custom models can then be used for transcribing speech. You can now train your models using AutoNLP, given you use text-based data. The service currently supports 5 languages.
Video Localisation
With this release, we launch a no-code Video Localization service that can overdub any video in your desired language. This is the tool you need to expand your audience by overdubbing your videos in multiple languages. Currently, we provide access to this service exclusively to Language Service Providers. Please feel free to reach out to us if interested. Our language support currently extends to 80+ languages.
Feel free to reach out to us or book a call directly if you’d like to talk with our team in more detail.
If you haven’t yet, sign-up on the NeuralSpace Platform to try and test it out by yourself! Get started with $200 worth of credits.
Be sure to check out our Documentation to read more about the NeuralSpace Platform and its different services.
Happy NLP!