-
Homepage
-
-
Inspiration
We love Manhwa, but often find ourselves short on time to sit down and read. After these comics are published, not all of them are adapted into anime, and even when they are, the process can take years—especially if the series spans multiple seasons. Rather than waiting or learning new languages, we decided to innovate something entirely new that the $8 billion manhwa/manga industry hasn't yet achieved at scale.
What it does
ComicSonic transforms Manhwa (Korean digital comics) into immersive audiobooks, giving users a hands-free, engaging storytelling experience. Using advanced audio processing, narration, and sound effects, ComicSonic brings characters and stories to life in vivid detail.
How we built it
We used Next.js for a responsive, intuitive frontend experience. Our backend, built with Python, processes uploaded Manhwa PDFs and leverages the Gemini API to accurately extract transcriptions. These transcriptions are then sent to Google Cloud Text-to-Speech, generating expressive, gender-specific audio. Finally, the audio files are seamlessly delivered to the frontend, where users can easily listen or download their favorite Manhwa audiobooks.
Challenges we ran into
Accurately capturing context and emotions from comic visuals was challenging. Fine-tuning our OCR and NLP models to precisely interpret the expressive, nuanced storytelling style of Manhwa required significant iterations and adjustments. Additionally, we faced challenges in maintaining the context of the content throughout longer documents, ensuring the narrative remained coherent and immersive.
Accomplishments that we're proud of
We're thrilled that ComicSonic delivers a genuinely immersive audio experience that closely matches the emotional depth of visual storytelling. Successfully automating the transition from visuals to expressive audio narratives within the hackathon timeframe felt especially rewarding.
What we learned
We deepened our expertise in audio generation, NLP, and the importance of emotional intelligence in storytelling. Additionally, working collaboratively taught us valuable lessons in agile problem-solving and rapid prototyping.
What's next for ComicSonic
Our next steps involve enhancing emotion detection accuracy, expanding language support, and adding user personalization features. Ultimately, we aim to partner with publishers to offer official audiobook versions of popular Manhwa series.
Log in or sign up for Devpost to join the conversation.