Parsa

open-ai-stylist. an open source ai stylist - hackathon winner

gaze-detection-video. Use the Moondream 2 model to detect faces and their gaze directions in videos.

video-understanding-engine. A powerful video summarization tool that utilizes Moondream alongside multiple AI models to provide comprehensive video understanding through audio transcription, intelligent frame selection, visual description, and content summarization.

fashn-tryon-extension. A Chrome extension that enables virtual fashion try-on and model swap using FASHN AI. Hover over fashion images on any website to: (1) try garments on your uploaded model images, or (2) swap the model’s identity while preserving the outfit.

open-ai-dress2impress. dress to impress, but with AI: featuring 3D avatars of the player and an AI opponent

open-impossible-video-generator. This project uses AI to generate "impossible" continuations of real videos. It takes a video input, analyzes its final frame, generates a creative scenario, and produces a continuation that defies reality in an entertaining way.

open-gtm-agents. open source gtm agents that grow your company | 🥇codex hackathon Apr 2026

uwRMP. uw rmp extension, integrate rmp directly into uw myplan website

moondream-local-vlm-nextjs-starter. Moondream Web Interface: A modern web interface for the Moondream vision language model, built with Next.js and FastAPI. This project provides a user-friendly way to interact with images using Moondream's vision-language capabilities.

omscs-turing-award-speaker-series. events page for 1st ever Turing award speaker series w/ 10+ Turing award winners & 2 noble laureate. hosted this with zach axel w/ direct support from founder of georgia tech OMSCS Zvi Galil.

object-detect-video. This tool uses Moondream 2B, a powerful yet lightweight vision-language model, to detect and redact objects from videos. Moondream can recognize a wide variety of objects, people, text, and more with high accuracy while being much smaller than most vision models.

promptable-content-moderation. Moderate any visual content that matches your prompt.

multimodal-image-similarity-search. An AI-powered image search application that combines text and visual similarity to help you find the right images. It understands both the visual content, semantics, and automatically captions all of your images. Offers dynamic filters for natural language filtering of images. Ex: "is this taken outdoors?" or "is there crime in this image?"

local-llm-coding-kit. A Windows-first setup for running Qwen3.6-35B-A3B locally with llama.cpp, then connecting coding CLIs such as Aider and opencode to the local OpenAI-compatible endpoint.

google-form-gen. A Next.js application that allows users to generate Google Forms programmatically using OAuth 2.0 authentication.

moondream. tiny vision language model

moondream-demo-video-ui-library. A specialized open source UI component library designed for video content creators, featuring animated components with chroma key backgrounds for easy integration into video productions.

moondream-benchmark. run to benchmark moondream 2bs performance