├── BenchmarkDLSoftware.md ├── Compression.md ├── History-CrossModal.md ├── History-Siamese.md ├── Homography.md ├── README.md ├── Stereo.md ├── Torch7-C-style-manual.md └── Torch7-Newbie.md /BenchmarkDLSoftware.md: -------------------------------------------------------------------------------- 1 | 2 | ## Papers 3 | - [Benchmarking State-of-the-Art Deep Learning Software Tools](http://arxiv.org/pdf/1608.07249v2.pdf), Arxiv 08/2016 [discussion] (https://news.ycombinator.com/item?id=12371508) 4 | 5 | ## Blogs 6 | - [Soumith Benchmark](https://github.com/soumith/convnet-benchmarks), Last updated 04/2015 7 | -------------------------------------------------------------------------------- /Compression.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | [TOWARDS THE LIMIT OF NETWORK QUANTIZATION](https://arxiv.org/pdf/1612.01543.pdf), ICLR 2017 under review. Sumsung US. 4 | 5 | [[Git](https://github.com/DeepScale/SqueezeNet/)][SqueezeDet: Unified, Small, Low Power Fully Convolutional Neural Networks for Real-Time Object Detection for Autonomous Driving] (https://arxiv.org/pdf/1612.01051.pdf) 6 | -------------------------------------------------------------------------------- /History-CrossModal.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | ## Thermal infrared feature 4 | - [[EOH](https://www.mathworks.com/matlabcentral/fileexchange/48200-ngunsu-matlab-eoh-sift)][Matlab] [Multispectral Image Feature Points] (http://www.mdpi.com/1424-8220/12/9/12661), Sensors 2012, 5 | - [EOH-Orientation] [Assigning Main Orientation to an EOH Descriptor on Multispectral Images] (http://www.ncbi.nlm.nih.gov/pubmed/26140348), Sensors 2015 6 | 7 | 8 | ## Cross-modal feature 9 | - [Multispectral Feature] [Multispectral Stereo Image Correspondence](http://www.cvc.uab.es/~asappa/publications/C__CAIP_2013_LNCS_8048_pp_217_224.pdf), CAIP 2013, UAB 10 | - [[MODS](http://cmp.felk.cvut.cz/wbs/index.html)][C++] [WxBS: Wide Baseline Stereo Generalizations](http://cmp.felk.cvut.cz/~matas/papers/mishkin-2015-WxBS-bmvc.pdf), BMVC 2015, Oxforad ETH 11 | - [[DASC](http://diml.yonsei.ac.kr/~srkim/DASC/)][Matlab] [DASC: Dense Adaptive Self-Correlation Descriptor for Multi-modal and Multi-spectral Correspondence] (http://www.cv-foundation.org/openaccess/content_cvpr_2015/papers/Kim_DASC_Dense_Adaptive_2015_CVPR_paper.pdf), CVPR 2016, Yonsei 12 | - [DeSCA] [Deep Self-Convolutional Activations Descriptor for Dense Cross-Modal Correspondence](http://arxiv.org/pdf/1603.06327v1.pdf), ECCV 2016, Yonsei 13 | 14 | ## Evaluation Report 15 | - [FIR-VS Feature] [A Novel SIFT-Like-Based Approach for FIR-VS Images Registration](http://www.cvc.uab.es/~asappa/publications/C__QIRT_2012_a.pdf), QIRT 2012, UAB 16 | -------------------------------------------------------------------------------- /History-Siamese.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | ### History 4 | - [Siamese Network] [Signature verification using a Siamese, NIPS 1994](https://www.google.co.kr/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0ahUKEwiw5dyq5Z3OAhWMp48KHY_ZCWUQFggiMAA&url=https%3A%2F%2Fpapers.nips.cc%2Fpaper%2F769-signature-verification-using-a-siamese-time-delay-neural-network.pdf&usg=AFQjCNGngnnSYTXzIRQzxrIfrmHwlzC_yQ&sig2=OTcUsyMU8EC1_h57xADdBA&cad=rjt) 5 | - [Siamese-Face Verification] [Learning a Similarity Metric Discriminatively, with Application to Face Verification](http://yann.lecun.com/exdb/publis/pdf/chopra-05.pdf), CVPR 2005 6 | - [[Triple Network](https://github.com/eladhoffer/TripletNet)] [Torch] [DEEP METRIC LEARNING USING TRIPLET NETWORK, ICLR 2015](http://arxiv.org/abs/1412.6622) [[ppt](http://tce.technion.ac.il/wp-content/uploads/sites/8/2016/01/Elad-Hofer.pdf)] 7 | 8 | 9 | ### Application 10 | 11 | - [Face Verification] [Learning a Similarity Metric Discriminatively, with Application to Face Verification](http://yann.lecun.com/exdb/publis/pdf/chopra-05.pdf), CVPR 2005 12 | - [Stereo][[MC-CNN](https://github.com/jzbontar/mc-cnn)] [Torch] [Stereo matching by training a convolutional neural network to compare image patches](http://arxiv.org/pdf/1510.05970v2.pdf), JMLR 2016 13 | - [Stereo][[MC-CNN](https://github.com/jzbontar/mc-cnn)] [Torch] [Computing the Stereo Matching Cost with a Convolutional Neural Network](http://arxiv.org/pdf/1510.05970v2.pdf), CVPR 2015 14 | - [[DeepDesc](http://cvlab.epfl.ch/research/detect/deepdescriptorlearning)][Torch] [Discriminative Learning of Deep Convolutional Feature Point Descriptors](https://infoscience.epfl.ch/record/213228/files/iccv-2015-deepdesc.pdf), ICCV 2015 15 | 16 | 17 | 18 | ### Tutorial & Practice 19 | - [Siamese Network：Architecture and Applications in Computer Vision](http://vision.ia.ac.cn/zh/senimar/reports/Siamese-Network-Architecture-and-Applications-in-Computer-Vision.pdf) 20 | - [Siamese Network Training with Caffe](http://caffe.berkeleyvision.org/gathered/examples/siamese.html) 21 | 22 | 23 | 24 | -------------------------------------------------------------------------------- /Homography.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | * [Deep Image Homography Estimation](https://pdfs.semanticscholar.org/91f1/05e040d33187ca4de7c0fcf7c6c942d8fe46.pdf) 5 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | 2 | ## [AI-Robotics Summer School Materials](http://www.kros.org/summerschool2016/02web01.php) 3 | - [Material] [FullBooks](http://multispectral.sejong.ac.kr/ykchoi/RSS2016/RSS2016-Print.pdf) : eBooks 4 | - [Material] [LocalFeatures](http://multispectral.sejong.ac.kr/ykchoi/RSS2016/Features.pdf) : From handcraft to deep learning 5 | - [Supp #1] [Binary Features](http://multispectral.sejong.ac.kr/ykchoi/RSS2016/RSS2016-BinaryFeatures.pdf) : BRIEF, ORB, BRISK, FREAK, BIO 6 | - [Supp #2] [Deep Features](http://multispectral.sejong.ac.kr/ykchoi/RSS2016/RSS2016-DeepFeatures.pdf) : DeepCompare, MatchNet, DeepDesc, PN-Net 7 | - Any questions? Email me or write issues. ykchoi@rcv.sejong.ac.kr / unizard@gmail.com 8 | 9 | ## [Feature's History](http://multispectral.kaist.ac.kr/ykchoi/RSS2016/History.png) 10 | 11 | 12 | ### Corner 13 | - [Harris][OpenCV] [Harris Corner Detector](http://faculty.cse.tamu.edu/jchai/csce641/harris_detector.pdf), 2004 [[ppt](http://www.cse.psu.edu/~rtc12/CSE486/lecture06.pdf)] York 14 | - [[FAST](http://www6.in.tum.de/Main/ResearchAgast)][OpenCV][C++] [FAST corner detection](https://arxiv.org/pdf/0810.2434.pdf), PAMI 2010 [[ppt](http://web.eecs.umich.edu/~silvio/teaching/EECS598_2010/slides/11_16_Hao.pdf)] Cambridge 15 | 16 | ### Conventional Feature 17 | - [[SIFT](http://docs.opencv.org/3.1.0/da/df5/tutorial_py_sift_intro.html#gsc.tab=0)][OpenCV][Vlfeat] [Object recognition from local scale-invariant features, ICCV 1999](http://www.cs.ubc.ca/~lowe/papers/iccv99.pdf), Lowe 18 | - [[SURF](http://docs.opencv.org/3.1.0/df/dd2/tutorial_py_surf_intro.html#gsc.tab=0)][OpenCV] [Speeded up robust features, ECCV 2006](http://www.vision.ee.ethz.ch/en/publications/papers/articles/eth_biwi_00517.pdf), ETH 19 | - [[DAISY](http://cvlab.epfl.ch/software/daisy)] [DAISY: An Efficient Dense Descriptor Applied to Wide Baseline Stereo, PAMI 2010](https://infoscience.epfl.ch/record/138785/files/tola_daisy_pami_1.pdf), EPFL 20 | 21 | ### Binary Feature 22 | - [[BRIEF] (http://docs.opencv.org/3.1.0/dc/d7d/tutorial_py_brief.html#gsc.tab=0)][OpenCV] [BRIEF: Binary Robust Independent Elementary Features, PAMI2012](https://infoscience.epfl.ch/record/167678/files/top.pdf), [ECCV2010](https://www.robots.ox.ac.uk/~vgg/rg/papers/CalonderLSF10.pdf), EPFL 23 | - [[ORB] (http://docs.opencv.org/3.1.0/d1/d89/tutorial_py_orb.html#gsc.tab=0)][OpenCV] [ORB: An efficient alternative to SIFT or SURF, ICCV 2011](http://www.willowgarage.com/sites/default/files/orb_final.pdf), Willowgarage 24 | - [[BRISK](http://docs.opencv.org/2.4/modules/features2d/doc/feature_detection_and_description.html#brisk)][OpenCV] [BRISK: Binary Robust Invariant Scalable Keypoints, ICCV 2011](https://www.robots.ox.ac.uk/~vgg/rg/papers/brisk.pdf), Oxford 25 | - [[FREAK](http://docs.opencv.org/2.4/modules/features2d/doc/feature_detection_and_description.html)][OpenCV] [FREAK: Fast Retina Keypoint, CVPR 2012](https://infoscience.epfl.ch/record/175537/files/2069.pdf), EPFL 26 | - [[AKAZE](https://github.com/pablofdezalc/akaze)][OpenCV] [Fast Explicit Diffusion for Accelerated Features in Nonlinear Scale Spaces, BMVC 2013](http://www.robesafe.com/personal/pablo.alcantarilla/papers/Alcantarilla13bmvc.pdf), GIT 27 | - [[BIO](https://sites.google.com/site/ykchoicv/feature_bio)][Matlab-mex] [Robust Binary Feature using Intensity Order, ACCV2014](https://www.dropbox.com/s/3glkyn0w0xxysij/ACCV2014_Robust%20Binary%20Feature%20Using%20Intensity%20Order.pdf?dl=0), KAIST 28 | 29 | ### Intensity based 30 | - [[LIOP](http://zhwang.me/publication/liop/index.html)][Matlab][Vlfeat] [Local Intensity Order Pattern for Feature Description, ICCV 2011](http://vision.ia.ac.cn/Students/bfan/LIOP-ICCV2011.pdf), CASIA 31 | - [[MROGH](http://www.openpr.org.cn/index.php/86-MROGH/View-details.html)][Matlab] [Aggregating Gradient Distributions into Intensity Orders, CVPR 2011](http://vision.ia.ac.cn/Students/bfan/BinFan_MROGH_CVPR11.pdf), CASIA 32 | 33 | ### Learning based 34 | - [[D-BREIF] (http://cvlab.epfl.ch/research/detect/dbrief)][Matlab] [Efficient Discriminative Projections for Compact Binary Descriptors, ECCV2012](http://cvlabwww.epfl.ch/~lepetit/papers/trzcinski_eccv12.pdf), EPFL 35 | - [[BinBOOST] (http://cvlab.epfl.ch/research/detect/binboost)] [[BinBoost: Boosting Binary Keypoint Descriptors](https://infoscience.epfl.ch/record/183635/files/top.pdf?version=1)], EPFL 36 | - [[VGG](http://www.robots.ox.ac.uk/~vgg/research/learn_desc/)] [Learning Local Feature Descriptors Using Convex Optimisation, PAMI 2014](http://www.robots.ox.ac.uk/~vgg/publications/2014/Simonyan14/simonyan13a.pdf), Oxford 37 | 38 | ### Neural Network based 39 | - [[DeepCompare](http://imagine.enpc.fr/~zagoruys/deepcompare.html)][Torch] [Learning to Compare Image Patches via Convolutional Neural Networks, CVPR 2015] (http://arxiv.org/pdf/1504.03641.pdf) 40 | - [[MatchNet](https://github.com/hanxf/matchnet)][Pycaffe] [MatchNet: Unifying Feature and Metric Learning for Patch-Based Matching, CVPR2015](http://www.cs.unc.edu/~xufeng/cs/papers/cvpr15-matchnet.pdf), Google 41 | - [[DeepDesc](http://cvlab.epfl.ch/research/detect/deepdescriptorlearning)][Torch] [Discriminative Learning of Deep Convolutional Feature Point Descriptors, ICCV 2015](https://infoscience.epfl.ch/record/213228/files/iccv-2015-deepdesc.pdf), [supp](http://hi.cs.waseda.ac.jp/~esimo/publications/supplemental/SimoSerraICCV2015_supp.pdf), [poster](http://hi.cs.waseda.ac.jp/~esimo/publications/posters/SimoSerraICCV2015_poster.pdf), EPFL 42 | - [[PN-Net](https://github.com/vbalnt/pnnet)][Torch] [PN-Net: Conjoined Triple Deep Network for Learning Local Image Descriptors, Arxiv 2016](http://arxiv.org/pdf/1601.05030v1.pdf), Imperial 43 | - [GLoss] [Learning Local Image Descriptors with Deep Siamese and Triplet Convolutional Networks by Minimising Global Loss Functions, CVPR 2016](http://arxiv.org/pdf/1512.09272v1.pdf), ACRV 44 | - [[LIFT](https://github.com/cvlab-epfl/LIFT)][Torch] [*LIFT: Learned Invariant Feature Transform, ECCV2016](http://arxiv.org/pdf/1603.09114v1.pdf), EPFL 45 | 46 | ### Evalation 47 | - [[Detector](http://www.robots.ox.ac.uk/~vgg/research/affine/evaluation.html#eval_soft)][Matlab] [A comparison of affine region detectors, IJCV 2005](http://www.robots.ox.ac.uk/~vgg/research/affine/det_eval_files/vibes_ijcv2004.pdf), Oxford, 48 | - [[Descriptor](http://www.robots.ox.ac.uk/~vgg/research/affine/desc_evaluation.html#code)][Matlab] [A performance evaluation of local descriptors, PAMI 2005](http://www.robots.ox.ac.uk/~vgg/research/affine/det_eval_files/mikolajczyk_pami2004.pdf), Oxford 49 | - [[Binary](http://cs.unc.edu/~jheinly/binary_descriptors.html)] [Comparative Evaluation of Binary Features, ECCV 2012](https://www.cs.unc.edu/~jheinly/publications/eccv2012-heinly.pdf), UNC 50 | - [[Orientation](https://github.com/cvlab-epfl/benchmark-orientation)] [Learning to Assign Orientations to Feature Points, CVPR2016](https://infoscience.epfl.ch/record/217982/files/top.pdf), EPFL 51 | - [Matching] [Evaluation of Local Detectors and Descriptors for Fast Feature Matching, ICPR] (http://www.miksik.co.uk/papers/miksik2012icpr.pdf), Krystian Mikolajczyk 52 | 53 | ### Evaluation Report 54 | - [A battle of three descriptors, 2012] (http://computer-vision-talks.com/2012-08-18-a-battle-of-three-descriptors-surf-freak-and-brisk/) 55 | - [Comparison of the OpenCV’s feature detection algorithms, 2011] (http://computer-vision-talks.com/2011-01-04-comparison-of-the-opencv-feature-detection-algorithms/) 56 | - [Comparison of the OpenCV’s feature detection algorithms – II, 2011] (http://computer-vision-talks.com/2011-07-13-comparison-of-the-opencv-feature-detection-algorithms/) 57 | - [Comparison of feature descriptors, 2011] (http://computer-vision-talks.com/2011-01-28-comparison-of-feature-descriptors/) 58 | 59 | 60 | ### Book 61 | - [Local Image Descriptor: Modern Approaches, 2015](http://www.springer.com/gb/book/9783662491713), Bin Fan 62 | - [Local Invariant Feature Detectors: A Survey, 2008](http://www.eng.auburn.edu/~troppel/courses/7970%202015A%20AdvMobRob%20sp15/literature/%5B2008%5D%20Local%20Invariant%20Feature%20Detectors-%20A%20Survey.pdf), Krystian Mikolajczyk 63 | 64 | 65 | 66 | ### Tutorial 67 | - [[Tutorial on Binary Descriptors](https://gilscvblog.com/2013/08/26/tutorial-on-binary-descriptors-part-1/)] [BRIEF](https://gilscvblog.com/2013/09/19/a-tutorial-on-binary-descriptors-part-2-the-brief-descriptor/), [ORB](https://gilscvblog.com/2013/10/04/a-tutorial-on-binary-descriptors-part-3-the-orb-descriptor/), [BRISK](https://gilscvblog.com/2013/11/08/a-tutorial-on-binary-descriptors-part-4-the-brisk-descriptor/), [FREAK](https://gilscvblog.com/2013/12/09/a-tutorial-on-binary-descriptors-part-5-the-freak-descriptor/), 2013 68 | 69 | -------------------------------------------------------------------------------- /Stereo.md: -------------------------------------------------------------------------------- 1 | 2 | Collection 3 | [Stereo] [Deep Stereo Matching with Dense CRF Priors](https://arxiv.org/pdf/1612.02401.pdf), 2016.12.08 4 | -------------------------------------------------------------------------------- /Torch7-C-style-manual.md: -------------------------------------------------------------------------------- 1 | 2 | ## Do it your self Torch7 - Basic 3 | 4 | ### Variables 5 | 6 | ### Loop control 7 | 8 | ### Operators 9 | 10 | - Tensor Resize 11 | 12 | inputs:resize(input:size()) 13 | 14 | - Tensor Resize & Copy 15 | 16 | inputs:resize(input:size()):copy(input) 17 | 18 | ### Functions 19 | 20 | ### Standard library 21 | 22 | ### Arrays 23 | 24 | ### Pointers 25 | 26 | ### Arrays and Pointers 27 | 28 | ### String function 29 | 30 | ### Structure 31 | 32 | ### Class 33 | 34 | 35 | ### 36 | nn.Sequential 37 | model:get(1) 38 | model:get(1):get(1) 39 | model:get(1):get(1):get(1) 40 | 41 | 42 | 43 | 44 | ## Do it your self Torch7 - Advanced 45 | 46 | ### Non-linearity 47 | - [ReLU] 48 | - [Leaky ReLU] 49 | - [PReLU] 50 | 51 | PReLU is not supported in CUNN, CUDNN. only support NN. 52 | 53 | example #1 54 | model:add(nn.PReLU()) 55 | 56 | example #2 57 | model:add(nn.PReLU(nil,nil,true)) 58 | 59 | - [RReLU] 60 | - 61 | 62 | 63 | 64 | 65 | 66 | 67 | 68 | 69 | 70 | -------------------------------------------------------------------------------- /Torch7-Newbie.md: -------------------------------------------------------------------------------- 1 | 2 | ### Torch7 for Newbie 3 | 4 | 5 | - [[Cheat Sheet](https://github.com/torch/torch7/wiki/Cheatsheet)] All about torch7 *: official documents* 6 | - [[60 Minutes!!](https://github.com/soumith/cvpr2015/blob/master/Deep%20Learning%20with%20Torch.ipynb)] Look at this basic pages 7 | - [[Tutorial & Demo](https://github.com/torch/torch7/wiki/Cheatsheet#tutorials-demos-by-category)] Look at the tutorial 8 | - [[Lecture](https://www.cs.ox.ac.uk/people/nando.defreitas/machinelearning/)] Oxford Computer Science - Machine Learning 2015, [Practicals](https://github.com/oxford-cs-ml-2015) 9 | 10 | - [[Torch with CaffeModel](http://gromit2.blogspot.kr/2016/07/neural-style-install-torch7-loadcaffe.html)] Torch with VGG19 11 | - [Luajit](http://luajit.org/luajit.html) Torch7 use Luajit which is a Just-In-Time Compilter for the Lua programming language. 12 | - [Lua 5.1 Manual](http://www.lua.org/manual/5.1/manual.html#2.2) [[pdf](http://multispectral.kaist.ac.kr/ykchoi/LuaManual.pdf)] 13 | - [Lua Sample Codes](http://lua-users.org/wiki/SampleCode) 14 | - [Torch site](http://torch.ch/docs/five-simple-examples.html) 15 | 16 | ### Community 17 | - [[Lua UserGroup](http://www.lua.org/community.html)] 18 | - [Torch7 UserGroup][[Open Chat](https://gitter.im/torch/torch7)] *: freely read/write question and answer* 19 | --------------------------------------------------------------------------------