streamdiffusion

C++ / CUDA implementation of StreamDiffusion

Real-time anime vision for VR headsets.

An interactive C++ drawing application that combines AI-powered visual completion through real-time canvas interpretation and image generation.

Python bindings for Spout2 GPU texture sharing on Windows. Wraps SpoutLibrary.dll via ctypes vtable dispatch — send and receive video frames with TouchDesigner, OBS, Resolume, and any Spout-aware application.

A real-time image generation service using StreamDiffusion, controllable via OSC (Open Sound Control) and communicating via Spout for GPU-accelerated texture sharing.

iOS app using Shazamkit to identify music and send info via OSC for auto-VJing.

Multisensory exhibition developed for the course Creative Programming and Computing at Politecnico di Milano, A.Y. 2025/26.

ATLAS — multi-team AI agent system with three-tier memory, critic-verified task loops, approval-gated host execution, and a live dashboard. Python + Next.js.

Protótipo de investigação: sinais do corpo lidos por câmara (rPPG, rosto, olhar) transformados em imagem e música geradas por IA, em tempo real, para ambientes sensoriais.

Realtime prompt-driven img2img on the Livepeer network: StreamDiffusion run unmodified behind the live runner

SurfaceShift — restyle any live screen with realtime diffusion (FLUX.2 cloud + on-device Apple Silicon fallback)

Streaming video diffusion on one AWS Trainium2 chip. StreamDiffusionV2’s causal Wan2.1 1.3B model ported to Neuron with a resident KV cache and a 4-NeuronCore layer pipeline: 5.6 → 38.4 FPS end to end, matching the CPU reference (cosine 0.99996). Plus a live SD-Turbo webcam restyle at 43.8 FPS.

Diffusion output as displacement, never as pixels — a height field and the record of how it settled

Generative AR Pet — a unique AI pet that lives in your camera, drawn in real-time by generative models