AccentAce

Inspiration

As a group of international students, we found ourselves grappling with the intricacies of pronunciation and accent improvement. Among the many linguistic hurdles we encountered, certain words such as 'chef' and 'genre' proved to be particularly elusive, often tripping us up and highlighting the need for focused attention on these nuances of language.

Despite the availability of apps like Duolingo and various websites that aid in reading and writing skills, a critical gap remains in assessing and improving pronunciation. This void inspired us to create AccentAce, an innovative application harnessing deep-learning networks to accurately evaluate and enhance a user's speaking proficiency.

What it does

AccentAce is a dynamic language-learning tool that caters to users of all proficiency levels. Within its user-friendly interface, you'll find three distinct modes to choose from: Easy, Medium, and Hard. In each mode, AccentAce harnesses the capabilities of the ChatGPT API to generate contextually relevant sentences that serve as practical exercises for pronunciation enhancement.

After you've been presented with a sentence, it's your turn to shine. You can record your voice, and this is where the power of our deep learning model comes into play. You can choose to refine your skills by tackling the same sentence again, using the feedback to perfect your pronunciation. Alternatively, if you're up for a fresh challenge, you can generate a new sentence, expanding your linguistic repertoire. AccentAce is more than just an app; it's your dedicated pronunciation coach, designed to unlock your full speaking potential and bolster your language skills.

How we built it

We harnessed the power of design with Figma to craft our website, ensuring an engaging and visually appealing user experience. Next, we seamlessly integrated the frontend with the backend using Taipy, creating a harmonious connection that ensures smooth and efficient operation.

But our innovation didn't stop there. We captured audio samples, and to bring them to life, we employed a sophisticated deep learning model, specifically a GRU (Gated Recurrent Unit). This advanced technology allowed us to classify and analyze the audio files with unparalleled accuracy and efficiency, opening up exciting possibilities for our project.

Challenges we ran into

Initially, our group had a lot of uncertainty as we grappled with the challenging task of choosing a direction. After much deliberation, we collectively decided to attempt React Native, drawn by the allure of venturing into uncharted territory. It was a bold choice, as none of us had prior experience with this technology, but our shared desire to explore something new propelled us forward.

However, as we delved deeper into React Native, we soon encountered formidable challenges that tested our collective resolve. The intricacies and complexities of this platform proved to be more demanding than any of us had initially anticipated. On the second day, we realized that the challenge was too much for our team, so we decided to pivot. In the end, we emerged from this experience not with regret, but with gratitude. We learned invaluable lessons in adaptability, problem-solving, and teamwork.

Accomplishments that we're proud of

We take immense pride in successfully completing our project on time, an achievement made even more significant due to the need to pivot from our original idea. This pivot tested our adaptability and resourcefulness as a team, highlighting the importance of staying agile in the face of evolving circumstances. What's equally rewarding is our unwavering commitment to giving our best effort throughout the journey, despite tackling an entirely new technology like React Native. The steep learning curve and challenges we encountered became opportunities for growth.

What we learned

At this event, we had the invaluable opportunity to acquire a diverse set of skills that enriched both our development capabilities and our ability to collaborate effectively in a team environment. Among the multitude of skills we cultivated, some of the standout ones included mastering Taipy, a valuable tool for bridging the gap between frontend and backend development. We delved into the intricacies of React Native and Expo, harnessing their potential to create powerful and user-friendly mobile applications. Additionally, our foray into the world of deep learning models not only expanded our technical expertise but also opened new doors to solving complex problems with AI-driven solutions. These newfound skills have not only bolstered our confidence as developers but also equipped us with the tools needed to tackle future challenges with proficiency and teamwork

What's next for Accent Ace

In the journey ahead, we have ambitious plans to further enhance our platform's capabilities and user experience. Our vision includes expanding the range of supported languages to cater to a broader audience, ensuring that learners worldwide can benefit from our pronunciation coaching. We're committed to implementing comprehensive progress-tracking features, allowing users to monitor their improvement over time and set personalized goals. To enhance the precision of our evaluations, we aim to refine our classification mechanisms, leveraging advanced AI techniques for even more accurate assessments. Additionally, we're dedicated to providing users with specific, actionable feedback on their pronunciation, empowering them with clear guidance on how to improve and refine their speech skills. These exciting developments are all part of our ongoing commitment to delivering a robust, user-centric language learning experience