DayBreak

超轻量级中文ocr，支持竖排文字识别, 支持ncnn、mnn、tnn推理 ( dbnet(1.8M) + crnn(2.5M) + anglenet(378KB)) 总模型仅4.7M

A lightweight OCR tool that reads Chinese (and other) text from images using three tiny AI models totalling under 5 MB, fast enough to run on phones and low-powered devices without a GPU.

yoloface大礼包 使用pytroch实现的基于yolov3的轻量级人脸检测（包含关键点）

Implementation Thundernet

加入关键点的darknet训练框架，轻量级的人脸检测，支持ncnn推理

the C++ version of solov2 with ncnn

centernet_mobilenetv2 inference by ncnn

thundernet ncnn

Implementation RefinedetLite

the C++ version of Seq2Seq with ncnn

the C++ version of Transformer with ncnn

train Snet(by thundernet) in imagenet

the C++ version of thundernet with ncnn

Video Question Answering

Object detection, 3D detection, and pose estimation using center point detection:

No description.

csv or xls input into hive

it's my personal website

A face detection algorithm

SOLOv2: Dynamic, Faster and Stronger, achives 39.5mAP on coco test-dev (36 epochs result)

SRMD super resolution implemented with ncnn library

The OCR approach is rephrased as Segmentation Transformer: https://arxiv.org/abs/1909.11065. This is an official implementation of semantic segmentation for HRNet. https://arxiv.org/abs/1908.07919

Android paddleocr demo infer by ncnn