Tag
#whisper
2 posts
Applied ai ml 10 min
Song to Screen: Brick-Built Music Videos on a DIY GPU Pipeline
How we turn a song into a finished brick-built animated music video with Suno, demucs, ComfyUI, and rented RunPod GPUs: real stages, real costs, real lessons.
video-generation comfyui runpod
Applied ai ml 15 min
Teaching AI to Watch Video: A Multi-Stage Pipeline
Video is the hardest media for AI to understand. Here's how I built SaySee — a pipeline that extracts frames, transcribes audio, and lets you search video by meaning.
video whisper embeddings