What’s your problem statement?

In the evolving landscape of content marketing, filmmakers and creators struggle to effectively engage diverse audiences due to the singular focus of traditional trailers on specific genres, which often overlook the multifaceted nature of their works. To address this challenge, we propose developing a platform that enables users to create multiple genre-specific trailers from a single movie or video. This solution will involve dynamic genre extraction, advanced AI-driven content segmentation, and a user-friendly interface that allows for easy uploads and trailer customization. By leveraging AI models to analyze key scenes and themes, the platform will generate engaging, cohesive trailers that highlight different narrative angles, ultimately enhancing the marketing potential of films and broadening their audience reach.

The problem SceneSplit solves

1. Multiple Genre-Specific Trailers: Create several trailers from one video to appeal to different audience interests.
2. Dynamic Genre Extraction: AI analyzes key scenes and themes to generate genre-appropriate content.
3. User-Friendly Interface: Easy uploads and customization without advanced editing skills required.
4. Time Efficiency: Automates the trailer creation process, saving valuable time for creators.
5. Enhanced Marketing Potential: Trailers highlight different narrative angles, broadening audience reach.
6. Increased Visibility: Tailored trailers boost visibility and connection with diverse viewer demographics.

Challenges we ran into

1. Versioning Challenges: We encountered difficulties managing different versions of our project due to a lack of clear version control practices, leading to confusion and delays.
2. Git LFS Limitations: The project included heavy files that exceeded Git's storage limits, resulting in issues with Git LFS. This hindered collaboration and slowed down our development process.
3. Model Hosting Constraints: We faced obstacles in hosting our AI model, as TensorFlow requires a well-configured cloud server. This limitation delayed our deployment timeline and required additional resources to address