NanoDet-Plus⚡Super fast and lightweight anchor-free object detection model. 🔥Only 980 KB(int8) / 1.8MB (fp16) and run 97FPS on cellphone🔥

🛠 A lite C++ toolkit of awesome AI models, support ONNXRuntime, MNN. Contains YOLOv5, YOLOv6, YOLOX, YOLOR, FaceDet, HeadSeg, HeadPose, Matting etc. Engine: ONNXRuntime, MNN.

🍅 Deploy ncnn on mobile phones. Support Android and iOS. 移动端ncnn部署，支持Android与iOS。

awesome AI models with NCNN, and how they were converted ✨✨✨

Deploy nanodet, the super fast and lightweight object detection, in your web browser with ncnn and webassembly

用opencv部署nanodet目标检测，包含C++和Python两种版本程序的实现

QuarkDet lightweight object detection in PyTorch .Real-Time Object Detection on Mobile Devices.

🥭 移动端MNN部署学习笔记。支持Android与iOS。

🍉 移动端TNN部署学习笔记，支持Android与iOS。

Tracking-by-Detection形式のMOT(Multi Object Tracking)について、 DetectionとTrackingの処理を分離して寄せ集めたフレームワーク(Tracking-by-Detection method MOT(Multi Object Tracking) is a framework that separates the processing of Detection and Tracking.)

nanodet int8 量化，实测推理2ms一帧！

手把手教你OpenVINO下部署NanoDet模型，intel i7-7700HQ CPU实测6ms一帧

NanoDet: Tiny Object Detection for TFJS and NodeJS

🍅🍅NanoDet、NanoDet-Plus with ONNXRuntime/MNN/TNN/NCNN C++. (https://github.com/DefTruth/lite.ai.toolkit)

A collection of some awesome public Anchor-Free object detection series projects.

NanoDet for a bare Raspberry Pi 4

NanoDetをGoogle Colaboratory上で訓練しONNX形式のファイルをエクスポートするサンプル(This is a sample to training NanoDet on Google Colaboratory and export a file in ONNX format)

本仓库在OpenVINO推理框架下部署Nanodet检测算法，并重写预处理和后处理部分，具有超高性能！让你在Intel CPU平台上的检测速度起飞！ 并基于NNCF和PPQ工具将模型量化(PTQ)至int8精度，推理速度更快！

NanoDet for Jetson Nano

NanoDetのPythonでのONNX推論サンプル