Inspiration
With the rise of AI and personalized content, we wanted to revolutionize how users consume news—making it more interactive, multilingual, and tailored to their interests through audio, video, and face recognition.
What it does
Our News Generator creates personalized news articles in audio and video formats, summarizes content in multiple languages, and uses face recognition to generate videos featuring specified celebrities related to the news.
How we built it
We used NLP for news summarization, text-to-speech (TTS) and video synthesis for multimedia generation, integrated translation APIs for multilingual support, and face recognition models to personalize content using celebrity visuals.
Challenges we ran into
Integrating various AI components like TTS, face recognition, and video generation while maintaining performance was challenging. Ensuring accurate celeb detection and syncing voice with generated visuals also took time.
Accomplishments that we're proud of
We successfully built an AI-powered system that offers a new way of news consumption—personalized, multimedia-rich, and multilingual. The face recognition to celeb-news-video pipeline was a standout feature.
What we learned
We deepened our understanding of multimedia generation, NLP, facial recognition, and system integration. We also learned how to manage performance and accuracy trade-offs in real-time content generation.
What's next for News Generator
We plan to add AR-based news experiences, real-time news feeds, emotion-driven summaries, and voice-interactive news playback. We also aim to improve celeb-video realism and expand to more global languages.
Log in or sign up for Devpost to join the conversation.