Banana Clips

🍌 BananaClips – Project Description BananaClips is an AI-powered consumer video editing tool built with Google Nano Banana.(Gemini 2.5 Flash Image model). It transforms raw videos into engaging, social-media ready clips by combining face detection, ASR transcription, and scene segmentation with Nano Banana 🔑 Key Innovations Nano Banana for Visual Understanding, face-aware processing: Detects and highlights speaking faces in frames Used for scene understanding so visual transitions align with speech cues. Speech + Scene Fusion Audio is transcribed via yt-dlp captions or Azure Batch Transcription. Speech timestamps are fused with Nano Banana scene detection to segment videos into natural storytelling clips. Each clip is subtitled with asr 🌟 Why It Stands Out Face detection → highlights speakers dynamically Scene detection → identifies natural cut points ASR alignment → syncs transcript with visual transformations