Study.md

Vision

Machine Learning

• Likelihood
  • Minimizing the Negative Log-Likelihood, in English : [링크]
• Kullback-Leibler Divergence
  • Kullback-Leibler Divergence Explained : [링크]
• Naive Bayes
  • Chapter 1 : Supervised Learning and Naive Bayes Classification — Part 1 (Theory) : [링크]
  • Naive Bayes - machine learning algorithm for classification problems : [링크]
• logistic regression model
  • HOW THE LOGISTIC REGRESSION MODEL WORKS, [링크]
• Machine learning
  • 머신러닝 모델링 알고리즘: [링크]
  • XOR problem theory : [링크]
  • Numerical Computation, [링크]
  • Introduction to Machine Learning, Amnon Shashua, [링크]
  • 가깝다, 멀다 - Distance / Metric의 개념, [링크]
• SVM
  • Chapter 2 : SVM (Support Vector Machine) — Theory, [링크]
• PCA
  • 차원축소 PCA : [링크]

Deep Learning

이론 설명

• Analysis of Dropout : [링크]
• Attention in Neural Networks and How to Use It : [링크]
• 루닛 논문 스터디 블로그, [링크]
• book
  • pdf, MIT Deep Learning Book in PDF format : [링크]
• ML Camp Jeju 2017 tutorial : [링크]
• 딥러닝 교육 자료 (Deep Learning Lecture) : [링크]
• A guide to convolution arithmetic for deep learning : [링크]
• 컨볼루션 이해하기, [링크]
• 카카오에서 12년간 발표 논문 6163건을 분석, [링크]
• [카카오AI리포트] 딥러닝 연구의 현재와 미래 (1/2), [링크]
• AI, ML, Deep Learning, [링크]
• A Neural Network in 11 lines of Python (Part 1), [링크]
• Applied Deep Learning Resources(정리페이지), [링크]
• Deep Learning Tutorial, dl_tutorials_10weeks [링크]
• Andrej Karpathy blog, Hacker's guide to Neural Networks, [링크]
• http://aikorea.org/cs231n/neural-networks-2-kr/
• Backpropagation
  • Gradient Descent Optimization Algorithms 정리 : [링크]
  • Gradient Descent Overview : [링크]
  • Derivation of Backpropagation : [링크]
  • Backpropagation 설명 예제와 함께 완전히 이해하기 : [링크]
  • Back-Propagation is very simple. Who made it Complicated ?, [링크]
  • 계산그래프로 역전파 이해하기, [링크]
  • Andrej Karpathy, Yes you should understand backprop : [링크]
    • 참고자료 : [링크]
    • TF backprop 구현: [링크]
  • A Derivation of Backpropagation in Matrix Form, [링크]
• Understanding CNNs : [링크]
• 딥러닝 관련 논문 정리 블로그 : [링크]
• 신경망 수학 연산 : [링크]
• python nn
  • en) http://iamtrask.github.io/2015/07/12/basic-python-network/
  • kr) http://ddanggle.github.io/ml/ai/cs/2016/07/16/11lines.html
• Softmax
  • Difference Between Softmax Function and Sigmoid Function, [링크]
• Collection of morning papers on Convolutional Neural Networks, [링크]
• Efficient Processing of Deep Neural Networks: A Tutorial and Survey, [링크]
• CNNs from different viewpoints, Prerequisite: Basic neural networks, [링크]
• batch normalization
  • ImageNet pre-trained models with batch normalization [링크]
    • https://github.com/cvjena/cnn-models/
    • http://www.inf-cv.uni-jena.de/Research/CNN+Models.html
  • 정리
    • Batch Normalization — What the hey? : [링크]
    • Batch Normalization 설명 및 구현
    • http://sanghyukchun.github.io/88/
    • http://funmv2013.blogspot.kr/2016/09/batch-normalization.html
    • Lab: Batchnormalization Layer, [링크]

Deep Reinforcement Learning

• StarCraft II - pysc2 Deep Reinforcement Learning Examples : [링크]

RNN

• NMT
  • Neural Machine Translation (seq2seq) Tutorial : [링크]
• RNN 튜토리얼, [링크]
• (한글 번역) Anyone Can Learn To Code an LSTM-RNN in Python (Part 1: RNN), [링크]
• 챗봇
  • DL Chatbot Seminar Day 01 : [링크]
  • tf, [링크]

CNN

• CNN을 모듈 방식의 관점, [링크]
• Convolutional Neural Networks (CNNs): An Illustrated Explanation, [링크]
• 개념이해(위한 소스)
  • opencv based, [링크]
  • https://github.com/xylcbd/EasyCNN
  • Convolutional Neural Networks (CNNs / ConvNets), [링크]
  • simple_cnn, [링크]
  • EasyCNN, [링크]
• Visualization
  • Picasso: A free open-source visualizer for Convolutional Neural Networks, [링크]

LIB

• PyTorch
  • PyTorchZeroToAll : [링크]
  • Kind_PyTorch_Tutorial, [링크]
  • PyTorch in 5 Minutes, [링크]
• Caffe C++
  • Blobs, 계층들, 그리고 망들 : Caffe 모델 분석 : [링크]
  • DeepLearning-Caffe-Nets-Layers-Blobs: [링크]
  • C++ Example 1. Hello Caffe, [링크]
  • Caffe c++ helloworld example with MemoryData input, [링크]
  • Making a Caffe Layer, [링크]
  • BVLC/caffe, Developing new layers, Developing new layers, [링크]
  • Data Parser in Caffe, [링크]
  • Simple Example: Sin Layer, Making Your First Layer, [링크]
  • Add new layer to Caffe, [링크]
• TensorFlow Tutorial
  • TensorFlow Datasets 및 Estimators를 소개합니다. [링크]
  • TensorFlow-Slim, [링크]
  • 텐서플로우(TensorFlow) 시작하기, [링크]
  • Tensorflow Tutorial 2: image classifier using convolutional neural network, [링크]
  • https://github.com/sherrym/tf-tutorial
  • RNN 모델
    • part 1 : [링크]
    • part 2 : [링크]
    • yutube : [링크]
  • https://github.com/golbin/TensorFlow-Tutorials
  • TensorFlow in 5 Minutes, [링크]
  • Getting started with Tensorflow, [링크]
  • https://www.youtube.com/watch?v=wuo4JdG3SvU)
  • 정리 : [링크]
  • TensorFlow Tutorial, [링크]
  • CS 20SI: Tensorflow for Deep Learning Research, [링크]
  • TensorFlow Basic Tutorial Labs, [링크]
  • tf-stanford-tutorials, [링크]
  • https://github.com/gicheonkang/TF-Tutorial
  • 기본강의, [링크]
  • Introduction to PyTorch and TensorFlow : [링크]
  • 코드 분석, [링크]

Relation Network

• A simple neural network module for relational reasoning : paper
• Relation Network : 링크
• Discovering objects and their relations from entangled scene representations : 링크
• A simple neural network module for relational reasoning 동영상 세미나

Classification

• model
  • Google Inception V3 : [링크]
  • ResNeXt(Aggregated Residual Transformations for Deep Neural Networks) : [링크]
• googlenet : https://github.com/lim0606/caffe-googlenet-bn
• Spatial Transformer Network
  • https://github.com/daerduoCarey/SpatialTransformerLayer
• Highway Networks
  • https://github.com/flukeskywalker/highway-networks/tree/master/examples/highways
  • http://people.idsia.ch/~rupesh/very_deep_learning/
• Specific Domain
  • nsfw
    • https://github.com/yahoo/open_nsfw

Compression Network

• SENet, Squeeze-and-Excitation Networks, [링크]
• ByteNet:Neural Machine Translation in Linear Time : [링크]
• SliceNet : depthwise separable convolution
  • Xception: Deep learning with depthwise separable convolutions
    • [링크]
    • https://github.com/yihui-he/Xception-caffe
  • tf function, [링크]
  • caffe, https://github.com/pby5/MobileNet
• Pruning Network
  • Pruning Neural Networks, [링크]
• Binary network
  • MobileNets
    • MobileNets: Open-Source Models for Efficient On-Device Vision :[링크]
    • MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications, [링크]
    • MobileNet-Caffe, [링크]
  • Ternary Weight Networks
    • caffe, https://github.com/fengfu-chris/caffe-twns
  • XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks : [링크]
    • http://www.erogol.com/xnor-net/
  • Quantized Convolutional Neural Networks for Mobile Devices, https://arxiv.org/abs/1512.06473
    • src) https://github.com/jiaxiang-wu/quantized-cnn
  • SqueezeNet
    • https://github.com/DT42/squeezenet_demo
    • Introducing SqueezeDet: low power fully convolutional neural network framework for autonomous driving, [링크]
    • https://github.com/songhan/SqueezeNet-Residual
    • caffe, [링크]
    • caffe, https://github.com/pmgysel/caffe/tree/master/models/SqueezeNet

Face

• facenet
  • https://github.com/davidsandberg/facenet
  • https://github.com/davidsandberg/facenet/wiki/Classifier-training-of-inception-resnet-v1
  • Triplet loss
    • https://github.com/hizhangp/triplet
    • https://github.com/luhaofang/tripletloss
• recongition & verification & feature
  • A Discriminative Feature Learning Approach for Deep Face Recognition
    • http://ydwen.github.io/
    • caffe src : [링크]
  • A Lightened CNN for Deep Face Representation
    • https://github.com/stoneyang/LCNN_TRAIN
  • SeetaFaceEngine
    • https://github.com/seetaface/SeetaFaceEngine
• facial landmark
  • https://github.com/luoyetx/deep-landmark
  • https://github.com/pminmin/caffe_landmark
  • MTCNN
    • https://github.com/DaFuCoding/MTCNN_Caffe
    • https://github.com/foreverYoungGitHub/MTCNN

Localization

• study
  • 루닛블로그, R-CNNs Tutorial, [링크]
  • Object Localization and Detection, [링크]
  • https://docs.google.com/presentation/d/1coDMqAVnorroIyrIloVf0hXDztyD6QVh5llvpX7VUDk/edit#slide=id.p
  • A Brief History of CNNs in Image Segmentation: From R-CNN to Mask R-CNN, [링크]
• adversarial-frcnn,
  • A-Fast-RCNN: Hard Positive Generation via Adversary for Object Detection, [링크]
  • src, [링크]
  • join GAN + fast rcnn
• faster rcnn
  • chainer src, 상세 그림, [링크]
  • caffe, resnet, [링크]
  • tf
    • Faster-RCNN_TF, [링크]
    • https://github.com/CharlesShang/TFFRCNN
  • Faster R-CNN, [링크]
  • squeezenet+faster r-cnn, hint, prototxt example, rbgirshick/py-faster-rcnn#345
• Yolo 2, YOLO9000: Better, Faster, Stronger : [링크]
  • caffe, https://github.com/gklz1982/caffe-yolov2
    • yolo1, https://github.com/yeahkun/caffe-yolo
  • caffe, yolo2, [링크]
• yolo
  • yolo, [링크]
  • Caffe Yolo
    • https://github.com/yeahkun/caffe-yolo
    • https://github.com/liuwei16/yolo-caffe
  • tf src
    • https://github.com/nilboy/tensorflow-yolo
    • https://github.com/gliese581gg/YOLO_tensorflow
    • https://github.com/hizhangp/yolo_tensorflow
    • https://github.com/dshahrokhian/YOLO_tensorflow
    • https://github.com/moontree/yolo_tensorflow
    • https://github.com/rokihi/ObjectDetector
    • https://github.com/iRiisH/yolo
    • Realtime iOS Object Detection with TensorFlow
      • https://github.com/yjmade/ios_camera_object_detection
      • https://github.com/Iflier/YOLO_TensorFlow_Application_Camera_RealTimeReco
    • https://github.com/Nielsyang/YOLOv1_tensorflow
    • v2, https://github.com/biyaa/yolov2_tensorflow
    • https://github.com/rhythm92/tf-yolo/tree/master/src
    • https://github.com/JindongJiang/yolo_tf
      • 비교적 깔끔한 소스, 그러나 실제 동작여부는 테스트하지 못함 
    • https://github.com/Bobeye/Poirot
  • study
    • https://docs.google.com/presentation/d/1kAa7NOamBt4calBU9iHgT8a86RRHz9Yz2oh4-GTdX6M/edit#slide=id.g15092aa245_0_184
    • https://pjreddie.com/publications/yolo/
    • video, https://www.youtube.com/watch?v=L0tzmv--CGY&feature=youtu.be
    • blog, https://curt-park.github.io/2017-03-26/yolo/
• PVANet
  • https://github.com/sanghoon/pva-faster-rcnn/tree/master/models/pvanet/example_train
  • https://github.com/sanghoon/pva-faster-rcnn
  • train : [링크]
  • PVANet-FACE, [링크]
• SSD
  • tf
    • https://github.com/balancap/SSD-Tensorflow
    • https://github.com/seann999/ssd_tensorflow
    • https://github.com/georgesung/ssd_tensorflow_traffic_sign_detection
    • https://github.com/ex4sperans/SSD

Segmentation

• Semantic Segmentation
  • A 2017 Guide to Semantic Segmentation with Deep Learning, [링크]
  • Image Segmentation using deconvolution layer in Tensorflow, [링크]
  • Deep Learning for Computer Vision - Going beyond image classification and regression, [링크]
  • PixelNet: Representation of the pixels, by the pixels, and for the pixels. [링크]
  • Deep learning for Earth Observation, [링크]
  • CRF-RNN for Semantic Image Segmentation, [링크]
  • Train CRF-RNN : [링크]
• Instance Segmentation
• fcn
  • study     * google slide : [링크]     * Fully convolutional networks for semantic segmentation. [링크]
  • tf version
    • https://github.com/yuxng/tensorflow-fcn
    • https://github.com/JindongJiang/FCN.tensorflow
  • caffe
    • https://github.com/shelhamer/fcn.berkeleyvision.org
    • https://github.com/bharatsingh430/py-R-FCN-multiGPU
    • https://github.com/Orpine/py-R-FCN
• Pyramid Scene Parsing Network
  • http://appsrv.cse.cuhk.edu.hk/~hszhao/projects/pspnet/index.html
  • https://github.com/CharlesShang/PSPNet

OCR

• Text Detection
  • Curated-scene-text-recognition-analysis : [링크]
  • SSD, [링크]
  • faster rcnn, [링크]
  • Detecting Text in Natural Image with Connectionist Text Proposal Network, [링크]
  • TextBoxes, based SSD , [링크]
  • Detecting Text in Natural Image with Connectionist Text Proposal Network, src, [링크]
• OCR
  • Image OCR:model architecture : [링크]

Generative Model

• Generative Model Adversarial Nets (GAN)
  • Deblur Photos Using Generic Pix2Pix, [링크]
  • The GAN Zoo, : [링크]
  • Collection of generative models in Tensorflow, [링크]
  • https://github.com/sanghoon/tf-exercise-gan : [링크]
  • CycleGAN : Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks - 컨셉 : [링크]
  • Chainer 기반의 다양한 GAN 을 구현한 라이브러리 : [링크]
  • (Pytorch를 사용한) 단 50줄로 코드로 짜보는 GAN : [링크]
  • Generative Models and GANs, [링크]
  • Introduction to domain adversarial training of neural network. [링크]
  • Continual Learning with Deep Generative Replay, [링크]
  • SK T-Brain Research, A Generative Model of People in Clothing, [링크]
  • Variants of GANs - Jaejun Yoo: [링크]
  • Understanding and Implementing CycleGAN in TensorFlow : [링크]
  • Deep generative model, [링크]
  • NIPS 2016, Generative Adversarial Networks
    • https://channel9.msdn.com/Events/Neural-Information-Processing-Systems-Conference/Neural-Information-Processing-Systems-Conference-NIPS-2016/Generative-Adversarial-Networks
    • https://arxiv.org/abs/1701.00160
  • Wasserstein GAN 수학 이해하기, [링크]
  • 아주 간단한 GAN 구현하기, [링크] 
  • Generative adversarial networks, [링크]
  • Deep Learning Research Review : [링크]
  • 초짜 대학원생의 입장에서 이해하는 Domain-Adversarial Training of Neural Networks (DANN)
    • http://jaejunyoo.blogspot.com/2017/01/domain-adversarial-training-of-neural-2.html
    • http://jaejunyoo.blogspot.com/2017/01/domain-adversarial-training-of-neural-3.html
    • http://jaejunyoo.blogspot.com/2017/02/deep-convolutional-gan-dcgan-1.html
    • http://jaejunyoo.blogspot.com/2017/02/deep-convolutional-gan-dcgan-2.html
  • InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets : [링크]
  • GAN 구현강좌 : [링크]
  • DiscoGAN
    • http://www.modulabs.co.kr/DeepLAB_library/12820
    • SK T-Brain Research, [링크]
• Variational autoencoder(VAE), [링크]
• On Face Segmentation, Face Swapping, and Face Perception
  • ttp://www.openu.ac.il/home/hassner/projects/faceswap/
  • http://www.creativeai.net/…/on-face-segmentation-face-swapp…
  • https://github.com/YuvalNirkin/face_swap
  • https://github.com/YuvalNirkin/face_segmentation
  • https://github.com/YuvalNirkin/face_video_segment

Style Transfer

• Style Transfer in Real-Time : [링크]
• texture Networks: Feed-forward Synthesis of Textures and Stylized Images
  • src, https://github.com/DmitryUlyanov/texture_nets
  • tf, https://github.com/tgyg-jegli/tf_texture_net

Similar Image

• Vector Similarity, "A Hybrid Geometric Approach for Measuring Similarity Level Among Documents and Document Clustering"
  • src, [링크]

CNN For Text Classification

• Text-CNN을 이용한 Sentiment 분설모델 구현 : [링크]
• UNDERSTANDING CONVOLUTIONAL NEURAL NETWORKS FOR NLP, [링크]
• http://www.wildml.com/2015/12/implementing-a-cnn-for-text-classification-in-tensorflow/)
• 합성곱 신경망(CNN) 딥러닝을 이용한 한국어 문장 분류 : [링크]
• tf code : [링크]
• Convolutional Neural Networks for Sentence Classification
• 자연어 처리 문제를 해결하는 CONVOLUTIONAL NEURAL NETWORKS 이해하기

Pose Estimation

• Realtime Multi-Person Pose Estimation: [링크]
• OpenPose: A Real-Time Multi-Person Keypoint Detection And Multi-Threading C++ Library, [링크]

SLAM

• ios orb slam
  • https://github.com/rhythm92/ORB_SLAM-IOS
  • https://github.com/egoist-sx/ORB_SLAM_iOS

Implementation

• Generative models
  • pixel-cnn++, Tensorflow : [링크]
  • Domain Transfer Network (DTN) : [링크]

Paper List

• http://www.arxiv-sanity.com/top?timefilter=alltime&vfilter=all
• http://trendingarxiv.smerity.com/
• conference
  • nips 2017, https://nips.cc/Conferences/2016/Schedule?type=Poster
  • nips list, https://papers.nips.cc/

Fast Search

• ANN
  • https://brunch.co.kr/@goodvc78/15
• Faiss by Facebook AI Research, https://github.com/facebookresearch/faiss
  • Billionscale similarity search with GPUs, https://arxiv.org/pdf/1702.08734.pdf

NLP

• 딥러닝 기반 자연어처리 기법의 최근 연구 동향 : [링크]
• word2vec
  • Sentiment analysis on forum articles using word2vec and Keras : [링크]
  • https://brunch.co.kr/@goodvc78/16

Math

• [Linear Algebra] Lecture 1, The Geometry of Linear Equations (1) : [링크]
• Essence of linear algebra : [링크]
• Linear algebra, https://www.slideshare.net/ssuser7e10e4/linear-algrbra

python

• numpy
  • https://www.youtube.com/watch?v=Fm1sWcdL_ak&t=298s
  • https://github.com/nsoojin/coursera-ml-py

Data Set

• Sketch RNN 기술에 이용된 벡터 이미지 데이터셋 : [링크]
• 정리 : [링크]
• Open Images dataset, https://github.com/openimages/dataset
• text detection, ocr : Scene Text Localization & Recognition Resources, [링크]
• e-Lab Video Data Set, e-VDS, [링크]