kenton-gmi/sakura-crossing
An explorable Japanese suburban railway-crossing neighbourhood on a small planet, rendered 3D-to-2D as a cel-shaded anime background. Three.js, no image assets.
07 / CREATIVE TOOLS
Explore open tools for images, sound, and moving pictures.
← All topicsAn explorable Japanese suburban railway-crossing neighbourhood on a small planet, rendered 3D-to-2D as a cel-shaded anime background. Three.js, no image assets.
Native support for JSX and TSX as a server and client rendered view engine for ASP.NET Core. Bringing the modern web to ASP.NET.
A macOS GUI app for designing and building custom DMG disk images
MAGI-2-preview: Scaling Video Generation Models Efficiently
Code and Model for Paper: Clarity Contrast and Similarity Selection for Multi-Focus Image Fusion
cadena: Stepwise CAD Reverse Engineering.
Open the project to explore the idea and the code.
100+ curated Seedance 2.5 prompts with real video previews, plus an installable Agent Skill that optimizes prompts, creates storyboards, and generates videos via Seedance models.
Spectrum-based acceleration for ComfyUI’s native MiniMax H3 audio-video model. Forecasts post-transformer features with Chebyshev ridge regression to skip selected transformer evaluations, with adaptive scheduling, sampler-aware safeguards, CPU/VRAM history storage, and native fallbacks.
A multi-hypervisor VM runtime for OCI images, supporting Cloud Hypervisor, Firecracker, QEMU, and Apple Virtualization.framework.
NV Broadcast — Unofficial NVIDIA Broadcast for Linux and other OS. AI-powered virtual camera with background removal, blur, replacement, video enhancement, and noise cancellation. GPU accelerated. Open source.
repo for QuerySplat.
Agentic CAD for 3D printing.
[CVPR 2026] Adaptive Spectral Feature Forecasting for Diffusion Sampling Acceleration
Open the project to explore the idea and the code.
Open the project to explore the idea and the code.
The most powerful local music generation model that outperforms almost all commercial alternatives, supporting Mac, AMD, Intel, and CUDA devices.
Docker images for LLM inference (SGLang + vLLM) on NVIDIA Blackwell GPUs (SM120, CUDA 13.2)
Agent skills that fact-check the internet: claim-by-claim verification with sources and a 0-10 BS score for any YouTube video, article, tweet, or PDF
Automatic Masking & Inpainting.
Open the project to explore the idea and the code.
An interactive 3D human anatomy explorer built using threejs with GPT 5.6 Sol
A free, open-source AI agent that runs entirely on your machine — controls your computer like you would, works across any app, and handles arbitrary tasks: writing code, analyzing images, organizing files, researching and building spreadsheets, whatever you throw at it. Unlike Claude Cowork, nothing leaves your device.
开箱即用的本地 JAV/视频 刮削、管理、播放软件,支持命令行一键安装和 docker 部署。只需简单添加目录,即可打造你的私人 JAV/视频 媒体库,带给你顶级的浏览体验,懒人必备。| Your local JAV/video manager.