Pinned Loading

1. Multimodal-RAG-Platform Multimodal-RAG-Platform Public

   A NotebookLM-inspired multimodal RAG platform for ingesting, encoding, and reasoning over text, images, and video with traceable, source-grounded generation.

   Python 4

2. CuVision-Engine CuVision-Engine Public

   low-latency Computer Vision framework written entirely in native C++/CUDA. Engineered for maximum throughput on NVIDIA hardware, CuVision-Engine provides optimized implementations for Classificatio…

   Cuda

3. PromptSeg-Lightweight PromptSeg-Lightweight Public

   A modular multimodal framework that generates object masks from text prompts. It uses a lightweight cross-modal decoder to fuse features from interchangeable vision and language backbones.

   Jupyter Notebook