An AI-Powered Speech Processing Toolkit and Open Source SOTA Pretrained Models, Supporting Speech Enhancement, Separation, and Target Speaker Extraction, etc.

Mooncake is the serving platform for Kimi, a leading LLM service provided by Moonshot AI.

A minimal and universal controller for FLUX.1.

Rust implementation of Ultralytics YOLOv8/v10 using ONNX (ort)

Fast ML inference & training for Rust with ONNX Runtime

Official repository of "SAMURAI: Adapting Segment Anything Model for Zero-Shot Visual Tracking with Motion-Aware Memory"

基于AI的图片/视频硬字幕去除、文本水印去除，无损分辨率生成去字幕、去水印后的图片/视频文件。无需申请第三方API，本地实现。AI-based tool for removing hard-coded subtitles and text-like watermarks from videos or Pictures.

Official Implementations for Paper - MagicQuill: An Intelligent Interactive Image Editing System

Try CoreML models on multiple images and videos easily and quickly

Header only C++ AES cipher library

Training-free Regional Prompting for Diffusion Transformers 🔥

free C++ class library of cryptographic schemes

Official inference framework for 1-bit LLMs

🚀 Next Generation AI One-Stop Internationalization Solution. 🚀 下一代 AI 一站式 B/C 端解决方案，支持 OpenAI，Midjourney，Claude，讯飞星火，Stable Diffusion，DALL·E，ChatGLM，通义千问，腾讯混元，360 智脑，百川 AI，火山方舟，新必应，Gemini，Moonshot …

SAM2Long: Enhancing SAM 2 for Long Video Segmentation with a Training-Free Memory Tree

Various AI scripts. Mostly Stable Diffusion stuff.

Official repository of In-Context LoRA for Diffusion Transformers

A simple screen parsing tool towards pure vision based GUI agent

Inference code for the paper "Spirit-LM Interleaved Spoken and Written Language Model".

DocLayout-YOLO: Enhancing Document Layout Analysis through Diverse Synthetic Data and Global-to-Local Adaptive Perception

Code for the paper Breaking reCAPTCHAv2 accepted at COMPSAC 2024

SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformer

ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator

Official code implementation of General OCR Theory: Towards OCR-2.0 via a Unified End-to-end Model

[NeurIPS 2024] VFIMamba: Video Frame Interpolation with State Space Models