:rotate(0)/a2u371eoh4tfzy28pql0qa54sbyw)
BRAGI
Research Question
What are the effects of augmented reality on people who struggle with language barriers in the near future?
BRAGI stands for Bilingual Reality Augmentation for Global Immersion. As an application for the Apple Vision Pro, it is designed to use language immersion with word recognition and translation to simulate an educational foreign language environment, not unlike Canada's L2 standard. BRAGI hopes to place a bet on the potentially massive AR boom within consumer hardware and aims to help with language barriers. Aside from direct language education, there could be potential for support with nonverbal autism, and providing future migrants with accessible language education when the hardware reaches commercial affordability. Further, by incorporating the AVP's object recognition, the application can harness familiar object salience and an adjustable spaced repetition to vary the level of immersion and difficulty concerning their ideal learning environment. There is also potential to implement OpenAI's whisper model, incorporating vocal translation and conversation prediction, allowing the user to practice conversations in varying languages. Another functionality could be the incorporation of Babbage/Davinci conversational models, allowing for response prediction and conversation support.
What is BRAGI?
Further Research
Immersion Education for the Millennium: What We Have Learned from 30 Years of Research on Second Language Immersion
This is a summary of the impact L2 immersion has on language education. This covers a 30-year study within Canada, supporting immersion within foreign languages as a learning method.
Further Research
Familiar Object Salience Affects Novel Word Learning
This is a review of the use of familiar object salience with preliminary language education
Pomper R, Saffran JR. Familiar Object Salience Affects Novel Word Learning. Child Dev. 2019 Mar;90(2):e246-e262. doi: 10.1111/cdev.13053. Epub 2018 Mar 7. PMID: 29512135; PMCID: PMC6128788.
Further Research
A trainable spaced repetition model for language learning
This is a study on the training of a HLR (half life regression) model for spaced reputation in a language learning environment. Specifically, this details how to modify a HLR model through the education process to dynamically adapt to the student.
https://aclanthology.org/P16-1174.pdf