MeowGPT team number: IM24

Inspiration

Establishing effective communication between cats and their humans has always been a fascinating challenge. We realised that, just like humans, cats also express their emotions through vocalisation. By understanding types of emotions behind the cats' meow and simulating the sound, we can create a two-way communication bridge between cats and humans.

What it does

MeowGPT is designed to detect and interpret the emotions behind a cat's meow, translating it into understandable feedback for the user. Whether a cat is feeling hungry, happy, or angry, our system provides the emotional states of the cat and can provide a response, enhancing the bond between the cat and its owner.

How we built it

We used convolutional neural network to process and analyse the cat's meow. We trained a model to classify the emotions of new, unheard meows. Auto encoding is also utilised in converting human speaks into cat language.

Challenges we ran into

The processing and feedback of data was crucial for the application's usability, requiring optimisation of the model.

Accomplishments that we're proud of

Achieving a certain accuracy rate in emotion detection, proving the viability of our deep learning approach. Creating a user interface that can create simulated cat sound contains emotional information.

What we learned

The usage of deep learning and neural network.

What's next for MeowGPT team number: IM24

We envision expanding MeowGPT's capabilities to identify more nuanced emotions and potentially detect health issues based on changes in vocal patterns. Additionally, we are trying to simulate more realistic cats' meow that can convey more message to the cats.