wangshujingscan / Nhyai
Licence: apache-2.0
AI智能审查,支持色情识别、暴恐识别、语言识别、敏感文字检测和视频检测等功能,以及各种OCR识别能力,如身份证、驾照、行驶证、营业执照、银行卡、手写体、车牌和名片识别等功能,可以访问网站体验功能。
Stars: ✭ 60
Programming Languages
python
139335 projects - #7 most used programming language
Projects that are alternatives of or similar to Nhyai
Fabrik
🏭 Collaboratively build, visualize, and design neural nets in browser
Stars: ✭ 1,081 (+1701.67%)
Mutual labels: caffe
Theano Kaldi Rnn
THEANO-KALDI-RNNs is a project implementing various Recurrent Neural Networks (RNNs) for RNN-HMM speech recognition. The Theano Code is coupled with the Kaldi decoder.
Stars: ✭ 31 (-48.33%)
Mutual labels: kaldi
Teacher Student Training
This repository stores the files used for my summer internship's work on "teacher-student learning", an experimental method for training deep neural networks using a trained teacher model.
Stars: ✭ 34 (-43.33%)
Mutual labels: caffe
My Very Deep Caffe
This is an implementation of very deep two stream CNNs for action recognition. The implementation is inspired by Wang et. al., https://github.com/yjxiong/caffe. Some improvements from Wang's implementation include reading videos from LDMB database, faster testing using LDMB interface. The aim is to work better with big dataset such as UCF101, HMDB51, Sports1M and ActivityNet easily.
Stars: ✭ 21 (-65%)
Mutual labels: caffe
Unreal caffe
Self Maintained Caffe. In this version Faster-RCNN, RFCN needs layer are fully supported!
Stars: ✭ 38 (-36.67%)
Mutual labels: caffe
Fashion Parsing
Repository of my fashion-parsing project. This project is put on hold since I am doing another project now, but will debug if bugs are reported.
Stars: ✭ 50 (-16.67%)
Mutual labels: caffe
Crnn
CRNN(Convolutional Recurrent Neural Network), with optional STN(Spatial Transformer Network), in Tensorflow, multi-gpu supported.
Stars: ✭ 34 (-43.33%)
Mutual labels: crnn
Keras Textclassification
中文长文本分类、短句子分类、多标签分类、两句子相似度(Chinese Text Classification of Keras NLP, multi-label classify, or sentence classify, long or short),字词句向量嵌入层(embeddings)和网络层(graph)构建基类,FastText,TextCNN,CharCNN,TextRNN, RCNN, DCNN, DPCNN, VDCNN, CRNN, Bert, Xlnet, Albert, Attention, DeepMoji, HAN, 胶囊网络-CapsuleNet, Transformer-encode, Seq2seq, SWEM, LEAM, TextGCN
Stars: ✭ 914 (+1423.33%)
Mutual labels: crnn
Bottom Up Attention
Bottom-up attention model for image captioning and VQA, based on Faster R-CNN and Visual Genome
Stars: ✭ 989 (+1548.33%)
Mutual labels: caffe
Convolutionalemotion
A deep convolutional neural network system for live emotion detection
Stars: ✭ 40 (-33.33%)
Mutual labels: caffe
Feathercnn
FeatherCNN is a high performance inference engine for convolutional neural networks.
Stars: ✭ 1,106 (+1743.33%)
Mutual labels: caffe
Genproto
An online GUI tool used to visualize prototxt and generate prototxt for caffe(current version).
Stars: ✭ 56 (-6.67%)
Mutual labels: caffe
nhyai 简介
基于深度学习的图像识别算法,海量大数据样本,准确识别图片和视频中的涉黄、 涉暴涉恐、政治敏感、民谣等内容,也能从美观和清晰等维度对图像进行筛选, 快速精准,解放审核人力。
演示站点: https://ai.hn-ssc.com/sindex
安装与使用
- Ubuntu18.04,Windows10虽然也能跑,但是环境整合会无比蛋疼
- python version >= 3.5.2
- pip install -r requirements.txt
- tensorflow,tensorflow-gpu version >= 1.13.1
- pip install torch torchvision
启用git lfs功能
- sudo apt-get install git-lfs
- git lfs install
前端打包
- cd /frontend
- npm run build
关于caffe安装
- sudo apt install caffe-cpu 或 sudo apt install caffe-cuda
- cp -r /usr/lib/python3/dist-packages/caffe /home/condas/envs/python3.6/lib/python3.6/
数据库割接:
-
python backend/manage.py makemigrations
-
python backend/manage.py migrate
后台运行:
- screen python backend/manage.py runserver 0.0.0.0:8000 --noreload --insecure
性能分析+后台运行:
- python backend/manage.py runprofileserver 0.0.0.0:8000 --noreload --insecure --kcachegrind --prof-path=/tmp/ai-profile-data
运行多任务:
- screen python backend/manage_task.py rqworker android ios default
创建用户Django
- python backend/manage_task.py createsuperuser
如遇到mysql安装错误,在ubuntu下执行如下命令:
- sudo apt-get install libmysqlclient-dev
项目截图
暴恐色情识别
通用识别
敏感文字过滤
视频检测
参考
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].