#IOS#YOLOv5 🚀 是在 COCO 数据集上预训练的一系列对象检测架构和模型，代表 Ultralytics 对未来视觉 AI 方法的开源研究，结合了经过数千小时研究和开发的经验教训和最佳实践。
#计算机科学#Netron是一种用于神经网络、深度学习和机器学习模型的可视化工具
#计算机科学#ONNX 是一种针对机器学习所设计的开放式的文件格式，用于存储训练好的模型。它使得不同的人工智能框架（如Pytorch、MXNet）可以采用相同格式存储模型数据并交互。
#计算机科学#ONNX Runtime：跨平台、高性能机器学习推理和训练加速器
#大语言模型#Unified framework for building enterprise RAG pipelines with small, specialized models
#计算机科学#🏄 Scalable embedding, reasoning, ranking for images and sentences with CLIP
#计算机科学#YOLOX is a high-performance anchor-free YOLO, exceeding yolov3~v5 with MegEngine, ONNX, TensorRT, ncnn, and OpenVINO supported. Documentation: https://yolox.readthedocs.io/
#下载器#A collection of pre-trained, state-of-the-art models in the ONNX format
#安卓#Open source real-time translation app for Android that runs locally
Go package for computer vision using OpenCV 4 and beyond. Includes support for DNN, CUDA, OpenCV Contrib, and OpenVINO.
Remove backgrounds from images directly in the browser environment with ease and no additional costs or privacy concerns. Explore an interactive demo.
#安卓#Sherpa-ONNX 是一个轻量级语音识别框架， 基于 Kaldi 和 onnxruntime，无需联网即可实现语音转文本、文本转语音、说话人分离以及语音活动检测(VAD)。支持嵌入式系统、安卓、iOS、鸿蒙系统、树莓派、RISC-V、x86_64 服务器、WebSocket 服务器 / 客户端，以及 C/C++、Python、Kotlin、C#、Go、NodeJS、Java、Swift、Dart、JavaScript、Flutter、Object Pascal、Lazarus、Rust 等编程语言。
#计算机科学#Setup and customize deep learning environment in seconds.
Silero VAD: pre-trained enterprise-grade Voice Activity Detector
MMdnn is a set of tools to help users inter-operate among different deep learning frameworks. E.g. model conversion and visualization. Convert models between Caffe, Keras, MXNet, Tensorflow, CNTK, PyT...
#大语言模型#Effortless data labeling with AI support from Segment Anything and other awesome models.
Silero Models: pre-trained speech-to-text, text-to-speech and text-enhancement models made embarrassingly simple