最近は化物語【ばけものがたり】を見た。かなり独特なアニメと思う。よく考えば、もう十二年前の作品だった。 化物語は名前の通り、怪異【かいい】の物語だ。阿良【あららぎ】々木 暦【こよみ】が解決した怪

Transformer Attention is all you need - Arxiv Feed Forward: Two fully connected layers (Linear) with a ReLU activation in between. Multi-head Attention: Attention Attention - Qiita Self-Attention GPT (Generative Pre-Training) GPT - paper BERT (Bidirectional Encoder Representations from

最近このゲームがめっちゃはやってるのて、調べだら、美しいゲーム画面に惹かれて、つい買っちゃった(・ω<) ﾃﾍﾍﾟﾛ。ちょっと遊んだら、やはりゲームデザイン的に面白いものもあるので、ちょっと分析

偶遇了这首治愈的歌。 Youtube player containerBack number - 水平線 一开始看到 MV 我还在想真是很省经费。但从主角收伞的那一刻就变得有趣（有趣？）起来了。一开始一直是被画面吸引，但注意到歌词后就有点被击中的感觉。

一开始我是被 PV 吸引了。有点意思，我想。女主有点酷；虽然 PV 完全意味不明，但是就是很戳。大概是因为色彩主题和几分前卫的内容。 第一集看下来，女主（似乎中文翻译应该是叫夏目？）果然很酷，而且穿

Tic-Tac-Toe Online Server Base on the Tic-Tac-Toe Game of CS188, Berkeley, I develop an online version of Tic-Tac-Toe. Now your agent can play with my agent online! I think it is a good way to check whether our agents are optimal

The Question Fish eating fruit on jisuanke Given an undirected acyclic graph G, all possible path P in the graph, calculate: The first taste In the contest, a handsome foriegn teammate conviced me that this problem can be solve using LCA. I tried. And it

Overview Regions with CNN features: Efficient Graph Based Image Segmentation use disjoint set to speed up merge operation Selective Search HOG (Histogram of Oriented Gradient) Multiple criterions (color, texture, size, shape) to merge regions AlexNet/VGG16 R-CNN Notice that many

Useful Materials Distinctive Image Features from Scale-Invariant Keypoints[1] by David G. Lowe. SIFT(Scale-Invariant Feature Transform)[2] on Towards Data Science. The SIFT (Scale Invariant Feature Transform) Detector and Descriptor[3]. Notes Uses DoG (Difference

Types of Noise Additive noise Additive noise is independent from image signal. The image g with nosie can be considered as the sum of ideal image f and noise n.[1] Multiplicative noise Multifplicative noise is often dependent on image signal. The relation of image and

Greek Alphabet Letter (Capital Case) Letter (Lower Case) Name English Equivalent alpha a beta b gamma g delta d epsilon e zeta z eta h theta th

Information Content where, I : the information content of X. An X with greater I value contains more information. P : the probability mass function. : b is the base of the logarithm used. Common values of b are 2 (bits),

AutoTag AutoTag is a program that generate tags for documents automatically. The main process includes: Participle (N-gram + lookup in dictionary) Generate bag-of-words for each document. Calculate term frequency and inverse document frequency. Pick top x words with greater

w3m: WWW wo Miru (c) Copyright Akinori ITO w3m is a pager with WWW capability. It IS a pager, but it can be used as a text-mode WWW browser. Keyboard Shortcuts Shortcut Action Level H Help Brower q Quit w3m

十年时间转瞬即逝，许多曾经深爱的事物都一一离去了，不免让人生出逝者如斯的感慨。 说来有些讽刺，作为一个计算机的学生，理应走在前沿，我却总怀念老式的交互和十年前的网络。 人生得意须尽欢，莫使金樽空

Friends' Blogs EquimA Kernel Implementation 長門有希An Outstanding ACMer Bug ZhangA Great Web Master Mr. ZhuA Dreamer Favorite Blogs CHARON(カロン)同人サークル「CHARON」の公式サイト 海底囚人海底囚人の趣味の個人サイトです

Permission Control for NTFS We often encounter the problem that to mount NTFS under Linux means no permission control. But that is not true. According to JanC's Answer on AskUbuntu: Contrary to what most people believe, NTFS is a POSIX-compatible filesystem, and it is

List # list filter table sudo iptables -L # list nat table sudo iptables -L -t nat Redirect # Redirect locally sudo iptables -A OUTPUT -t nat -p tcp --src 127.0.0.1 --dport 80 -j REDIRECT --to-port 8080

Major and Minor Step The step between each note below is a half: C C#(Db) D D#(Eb) E F F#(Gb) G G#(Ab) A A#(Bb) B C Steps in Major and Minor S: a half step T: a

Substitution, dirname, basename and suffix Substitution can be used to get path and short filename. filename=a/b/c/name.file echo ${filename#*/} # b/c/name.file echo ${filename##*/} # name.file echo ${filename%/*

Input and Output $ ffmpeg -i input.mp4 output.mp4 Cutting Video or Audio Begin and length: $ ffmpeg -i input.mp4 -ss 00:00:10 -t 01:00:10 output.mp4 Beginning time and ending time: $ ffmpeg -i input.mp4 -ss 01

“另外”司机对着后视镜说,“有一件事想請您记住: 事物往往和外表不一样。" 事物往往和外表不一样。青豆在脑中重复了一遍, 微颦眉尖。“什么意思?" 司机字斟句酌地说:“就是说,现在您要去做一

0. install GPU drivers sudo add-apt-repository ppa:graphics-drivers/ppa sudo apt update sudo apt install nvidia-390 1. install cuda tookit and cudnn SDK (and CUPTI) Following the instructions at Installing Tensorflow on Ubuntu. # Adds NVIDIA package repository. sudo apt-key

This article is my reflection on my previous work FaceLock, a project to recognize user's face and lock the computer if the user doesn't present in a certain time. CNN is used to recognize different faces. I watch the Coursera course Convolutional Neural Networks by

This article is about some squashing functions of deep learning, including Softmax Function, Sigmoid Function, and Hyperbolic Functions. All of these three functions are used to squash value to a certain range. Softmax Function Softmax Function: A generalization of the logistic function that "squashes" a

This article is my learning note of the coursera course Sequence Models by Andrew Yan-Tak Ng. There are two typical RNN units of the hidden layers of the RNN according to Andrew Ng. One is GRN (Gated Recurrent Unit), the other is LSTM (Long Short

Transformation Defines a struct ObjectProperty as follow: property type comment rotation vec3 The object's rotation in its object center. scale vec4 / vec3 The object's scale level in its object center. translation vec3 The object's relative position to the

Time Complexity With partition, the elements compared each time are as follow: n, n/2, n/2, n/4, n/4, n/4, n/4, ..., 1, 1, ..., 1 (n ones) And their sum is: n, n,

以下皆为在开发我的 C++ 库 donnylib 的时候遇到的问题。 （一天遇到这么多 weird problems 也是很走运了） donnylib : https://github.com/Donny-Hikari/donnylib Weird Template Subclass 如下的程序会导致无法自动推断模板类型。（编译环境

Python 实现: AdaBoost - Donny-Hikari - Github Introduction AdaBoost 是 Adaptive Boosting 的简称。 Boosting 是一种 Ensemble Learning 方法。 其他的 Ensemble Learning 方法还有 Bagging, Stacking 等。 Bagging, Boosting, Stacking 的区别如下： Bagging: Equal weight voting. Trains

Android Library To build a android library project, simply change the following line in build.gradle (Module): apply plugin: 'com.android.application' to: apply plugin: 'com.android.library' To pack the library into a jar file, add these lines to build.gradle

This is a learning note of Logistic Regression of Machine Learning by Andrew Ng on Coursera. Hypothesis Representation Uses the "Sigmoid Function," also called the "Logistic Function": Which turn linear regression into classification. Sigmoid function looks like this: give us the probability that

Concept 上下界网络流 顾名思义，比普通的网络流多出了下界。 考虑将下界的流量限制进行转化。将原来的边(e, v1->v2)拆边，拆出一条上界为 max-min 的边(e1)，和一条上界为 min 的边(e2, 虚拟存在

Problem 合怪升级，每只怪有他的等级和攻击力，以及是否为 tuner monster. 合怪必须一只 tuner monster 与一只 non-tuner monster 合，并且他们的等级之和必须等于合成的怪。只有特定等级和攻击力的怪可以被合成。某些合成怪必须

Introduction 最精彩的证明是自动证明。今天在知乎上看到的一位大佬对莫比乌斯反演的证明让我体会到了数论的神奇。坚定了我投靠数学的决心。 Mobius Inversion Formula 若存在 则有 反之亦然。 Basic Concept 1. Convolution /

Problem 给定数字串 S，求如下表达式的值： Solution 用前缀和转化该式。令 , 则 . 原式转化为： 二项展开，得： 将内外求和对调，使得可以预处理前缀和的和的p次方。式子化为：

最近（上个月）想研究一下人脸识别。人脸检测。找了下网上的资料，有不少都是介绍怎么用现成的模块识别的。但是我想了解的是用神经网络进行人脸识别，并且希望能够更多地接触神经网络。于是往基于TensorFlow框架的人脸