One strength of Auris is its adaptiveness. The tool automatically and precisely detects each learner's needs and creates a personalized training program. Through playing a fun game, learners train their brains to identify and discriminate challenging sounds of their second language.

The previous adaptive algorithm in Auris was designed to start slow, with just one exercise type, and then gradually add other types of exercises as the system builds a more robust learner model. The updated algorithm improves learner engagement without compromising effectiveness. Starting today, learners will see a wider variety of exercises in Auris from the start!

In Auris, the words that students hear during training are not recorded by a live person, but instead computationally synthesized using text-to-speech technology. This approach helps control for the acoustic parameters of the stimuli to avoid any irrelevant cues (such as background noises and slight variations in the intonation) that might distract learners and slow down their progress.

In this upgrade, we have incorporated newer text-to-speech technology that uses artificial neural networks to produce higher-quality output. We have also followed a rigorous stimulus screening process to make sure that no stimuli that might confuse learners are ever presented in the training process.

To diagnose each learner's unique needs, Auris uses an implicit assessment procedure: Instead of administering tests, the system continually monitors the learner's performance in training activities and makes inferences about their progress and challenges. The benefit of the implicit assessment approach is that it eliminates the stress of overt testing and makes learning more fun.

With the new update, we added an assessment of how well the learner can generalize their knowledge of sounds to new voices and new phonetic contexts. This generalization ability is the ultimate goal of sound learning and is routinely measured in laboratory research; however, such tests do not tend to be very engaging, and we have struggled for a long time with how we could incorporate them into Auris. Now, we believe, we are giving your students the best of both worlds!

Learners and instructors do not need to change how they work with Auris. The new assessment protocol starts automatically at carefully chosen moments in time to fine-tune each learner's individualized training program. Learners might notice two small changes to their experience:

We worked hard to increase the range of English sounds that are covered by Auris. This meant identifying pedagogically relevant sound pairs, preparing and screening stimuli, and running multiple internal tests to make sure that the new material works well with the rest of the system.

After a successful pilot study in the spring, we feel that we're ready to release a French version of Auris as a public beta. Currently, the French system only focuses on vowels, which are known to be the most challenging sounds for learners of French. Please contact us if you are interested in trying out Auris with your French language students.

We are excited about these updates, and we look forward to seeing your students benefit from them!

P. S. By the way, we've also been working on a beta version of Lexis for Spanish. If you would like to know more about it, just drop us a line.