Traditional music visualization relies heavily on predefined procedural mappings. This project explored generative video diffusion models to synthesize dynamic, context-aware visual accompaniments for musical audio.

We engineered a pipeline mapping acoustic features extracted via the Spotify API directly into the latent space coordinates of video diffusion architectures. To validate the visual coherence and aesthetic appeal, we designed a custom quantitative evaluation metric and conducted a user study with over 50 participants.

Quantum processors in the NISQ era are highly susceptible to environmental decoherence and gate errors. Gauging algorithm resilience under these conditions is key to designing noise-tolerant quantum systems.

This thesis evaluated Grover's amplitude amplification algorithm run on physical IBM Q processors. We tracked error accumulation across multiple amplification steps, mapping out exact noise thresholds and comparing the performance differences across layout topologies and qubit designs.