Storiboard(UK)

Key Features:

1. Transcription and Translation:

o Utilized the OpenAI Whisper model (large v2) to generate accurate transcriptions for videos in multiple languages. This model was chosen for its robustness in handling diverse linguistic nuances and providing high quality text outputs.
o Enabled translation capabilities, allowing users to transcribe videos from different languages seamlessly.

2. Video Cleaning and Editing:

o Implemented functionalities to detect and remove silent parts and unwanted filters from the videos, ensuring a cleaner final product.
o Provided tools to edit the transcription, which in turn allowed precise editing of the video content, making the process intuitive and user friendly.

3. Sentiment Analysis and Topic Generation:

o Integrated sentiment analysis to evaluate the emotional tone of the video content. This feature helped users understand the overall sentiment conveyed in the video.
o Developed a topic generation module to identify and summarize the main topics discussed within the video. This was achieved using advanced natural language processing techniques, ensuring accurate and relevant topic extraction.

Technology Stack

• OpenAI Whisper Model: For high-quality transcription and translation.
• Stable Whisper: Used for generating word-level timestamps, which are crucial for precise video editing.
• MoviePy: A Python library for video editing, enabling functionalities like cropping, clipping, and concatenating video segments.