投稿時間:2022-07-04 05:21:04 RSSフィード2022-07-04 05:00 分まとめ(27件)

カテゴリー等	サイト名等	記事タイトル・トレンドワード等	リンクURL	頻出ワード・要約等/検索ボリューム	登録日
AWS	AWS Podcast	#534: Mental Models for Technologists to Become Business Leaders	https://aws.amazon.com/podcasts/aws-podcast/#534	Mental Models for Technologists to Become Business LeadersIn this episode Simon sits down in the studio with Iain Rouse GM for AWS Public Sector across Australia New Zealand and Oceania The two have an informative chat about how technologists can develop skills to become business leaders how to create transparency through ceremony within your team and making tough decisions to ensure your business stays agile and resilient Leave us feedback	2022-07-03 19:32:54
python	Pythonタグが付けられた新着投稿 - Qiita	SQLAlchemy modelにsetterとgetterを設定する方法	https://qiita.com/GuriTech/items/eb1b4631e90e0b33a746	attribute	2022-07-04 04:44:25
海外TECH	MakeUseOf	How to Capture Network Traffic in Linux With tcpdump	https://www.makeuseof.com/tcpdump-linux-network-traffic-capture/	linux	2022-07-03 19:30:14
海外TECH	DEV Community	How do Convolutional Neural Networks work?	https://dev.to/ashwinscode/how-do-convolutional-neural-networks-work-2c23	How do Convolutional Neural Networks work Facial Recognition Object Detection and automated cancer detection how do these all work At the core of many state of the art image processing models lies Convolutional networks which use what are known as convolutions to accurately capture the features of an image Not only can CNNs be used for image processing they can also be handy in NLP tasks often serving as an alternative to recurrent neural nets due to the fact they can be trained much faster This article however will focus mainly on CNNs with an image processing task in mind What are Convolutions At the core of CNNs is an operation known as a convolution Each convolutional layer performs this operation A convolution uses a filter and passes this filter over a given D array of numbers performing a weighted sum using the numbers the filter passes over to create a feature map For example As the diagram shows a x filter and a x D array of numbers have been setup Now let s do the first step of passing this filter over the D array The filter is applied to the first x section of the D array performing a weighted sum of the numbers it covers In this case the weighted sum is The result of this sum is then added into the feature map Stride length specifies how many columns the filter should move across after doing a weighted sum In our example we will set the stride length to This means the following occurs This process repeats until the filter has passed through the whole D array forming a full feature map In terms of image processing images can be represented as a D array of numbers with each element representing an image pixel After a convolution is applied the feature map if a useful filter is used should represent a meaningful feature of the image Exploring convolution filtersThere are some commonly known filters which result in interesting feature maps Below is a Gaussian blur filter left begin matrix amp amp amp amp amp amp end matrix right ⎝⎛​​​​⎠⎞​and here is a short script which applies this convolves an image using this filter and shows the result from matplotlib import imagefrom matplotlib import pyplotfrom scipy ndimage import convolveimport numpy as npdef rgbgray rgb r g b rgb rgb rgb gray r g b return grayimage rgbgray image imread city jpg converts to grayscale for simpler convolvingblur kernel np array pyplot imshow image cmap gray vmin np min image vmax np max image pyplot show convolved convolve image blur kernel pyplot imshow convolved cmap gray vmin np min image vmax np max image pyplot show Using this script if we inputted this image the convolution s feature map would look like this As you can see this particular filter blurs an image its given Another well known filter is this edge detection filter left begin matrix amp amp amp amp amp amp end matrix right ⎝⎛​​ ​​⎠⎞​Here is the same script but the kernel filter has been changedfrom matplotlib import imagefrom matplotlib import pyplotfrom scipy ndimage import convolveimport numpy as npdef rgbgray rgb r g b rgb rgb rgb gray r g b return grayimage rgbgray image imread city jpg edge kernel np array pyplot imshow image cmap gray vmin np min image vmax np max image pyplot show convolved convolve image edge kernel pyplot imshow convolved cmap gray vmin np min image vmax np max image pyplot show and here is the resulting image As shown this particular filter detects any edges in the input image Hopefully you can see that filters are used to pick out specific features in an image and we can use different filters to pick out different features we would like to see How are Convolutions used in CNNs Learning FiltersObviously manually specifying what convolution filters the network should use would most likely yield poor results CNNs initially start with random filters and adjust them during the training process which eventually results in the optimal filters which pick out specific features in the input which would not be obvious to a human eye Multiple FiltersA convolutional layer does not just use one filter they learn many filters in parallel usually from to filters This allows the network to look at the input from different perspectives and pick up very specific details within the image e g certain curve shapes Multiple ChannelsColoured images consists of channels usually having one channel for each RGB value Images will now have a D shape meaning the filter must also have a D shape e g having a xx filter instead of a x The mechanism is the same instead the weighted sum happens across multiple channels instead of one Working with multiple channels allows the network to work with more data from images and therefore develop more complex understandings of them PoolingAlong with Convolutional layers pooling layers are commonly used There are main types of Pooling Average Pooling Max Pooling Global PoolingPooling is a type of dimensional reduction where a window slides across a D array of numbers and the numbers in within this window is reduced to a single number Average PoolingThe average of all numbers within the window is taken and kept into the new resulting D array Max PoolingThe largest number within the window is taken and kept into the new resulting D array Global PoolingIf you wanted to pool a D array meaning the input has channels you could either Global Max Pool it or Global Average Pool it They reduce the channels into channel using the maximum channel values or average channel values respectively Thank YouThank you for reading and hopefully you understand the theory behind CNNs If you are interested in implementing a CNN here is how you could do it in Tensorflow	2022-07-03 19:35:41
Apple	AppleInsider - Frontpage News	New Apple Music student subscribers can get free Beats Flex	https://appleinsider.com/articles/22/07/03/new-apple-music-student-subscribers-can-get-free-beats-flex?utm_medium=rss	New Apple Music student subscribers can get free Beats FlexAlongside its existing back to school offers Apple is providing new subscribers of its Apple Music Student plan a free pair of Beats Flex earphones Available to new accounts subscribing to Apple Music on the cheaper Student plan starting from June the offer provides users in the United States and the United Kingdom free Beats Flex earphones According to the terms and conditions of the offer it applies only to students in the two countries who haven t previously subscribed to any Apple Music tier Qualifying for an Apple Music Student subscription requires studying a bachelor degree post graduate degree or equivalent higher education course and be verified as a student either through Apple or a supported third party verification partner Read more	2022-07-03 19:42:23
Apple	AppleInsider - Frontpage News	Deals: Apple's 14-inch MacBook Pro is still on sale for $1,799 ($200 off), M1 13-inch drops to $1,049	https://appleinsider.com/articles/22/07/03/deals-apples-14-inch-macbook-pro-is-still-on-sale-for-1799-200-off-m1-13-inch-drops-to-1049?utm_medium=rss	Deals Apple x s inch MacBook Pro is still on sale for off M inch drops to The inch MacBook Pro promotion at leading Apple resellers is still one of the best th of July deals available this holiday weekend Best Buy has also dropped the M inch MacBook Pro to while supplies last Apple s inch MacBook Pro and MacBook Pro are discounted for the th of July inch MacBook Pro price war Read more	2022-07-03 19:22:56
海外科学	NYT > Science	James Bardeen, an Expert on Unraveling Einstein’s Equations, Dies at 83	https://www.nytimes.com/2022/07/03/science/space/james-bardeen-an-expert-on-unraveling-einsteins-equations-dies-at-83.html	James Bardeen an Expert on Unraveling Einstein s Equations Dies at A scion of a renowned family of physicist he helped set the stage for what has been called the golden age of black hole astrophysics	2022-07-03 19:57:52
ニュース	BBC News - Home	Several killed in Copenhagen shopping mall shooting	https://www.bbc.co.uk/news/world-europe-62030919?at_medium=RSS&at_campaign=KARANGA	fields	2022-07-03 19:14:53
ニュース	BBC News - Home	Army's YouTube and Twitter accounts hacked	https://www.bbc.co.uk/news/uk-62030644?at_medium=RSS&at_campaign=KARANGA	british	2022-07-03 19:34:52
ニュース	BBC News - Home	Wimbledon 2022: Cameron Norrie beats Tommy Paul to reach quarter-finals at All England Club	https://www.bbc.co.uk/sport/tennis/62030683?at_medium=RSS&at_campaign=KARANGA	Wimbledon Cameron Norrie beats Tommy Paul to reach quarter finals at All England ClubCameron Norrie maintains the British challenge in the Wimbledon singles by beating American Tommy Paul to reach his first Grand Slam quarter final	2022-07-03 19:30:44
ニュース	BBC News - Home	India v England: Day three highlights - Bairstow makes another century but India remain in control	https://www.bbc.co.uk/sport/av/cricket/62031173?at_medium=RSS&at_campaign=KARANGA	jonny	2022-07-03 19:01:52
ニュース	BBC News - Home	British Grand Prix: Lewis Hamilton backs protests but Mercedes question methods	https://www.bbc.co.uk/sport/formula1/62030105?at_medium=RSS&at_campaign=KARANGA	British Grand Prix Lewis Hamilton backs protests but Mercedes question methodsLeading drivers said they backed the cause of protesters who demonstrated on environmental issues after a track invasion at the British Grand Prix	2022-07-03 19:21:11
ニュース	BBC News - Home	England v India: Jonny Bairstow touching new heights in Ben Stokes era	https://www.bbc.co.uk/sport/cricket/62030245?at_medium=RSS&at_campaign=KARANGA	England v India Jonny Bairstow touching new heights in Ben Stokes eraJonny Bairstow s stellar form is a product of feeling valued supported and important in the England team says chief cricket writer Stephan Shemilt	2022-07-03 19:40:11
ビジネス	ダイヤモンド・オンライン - 新着記事	生保業界が「金融DX最後発」に陥る理由、捨てるに捨てられない超レガシーシステムの呪縛 - 金融DX大戦	https://diamond.jp/articles/-/305400	生命保険	2022-07-04 05:00:00
ビジネス	ダイヤモンド・オンライン - 新着記事	INPEX・ENEOS・出光興産「原油高＋α」で超好決算！売上高倍増、営業利益3倍… - ダイヤモンド 決算報	https://diamond.jp/articles/-/305824	INPEX・ENEOS・出光興産「原油高α」で超好決算売上高倍増、営業利益倍…ダイヤモンド決算報コロナ禍が落ち着き始めたことで、市況も少しずつ回復しつつある。	2022-07-04 04:55:00
ビジネス	ダイヤモンド・オンライン - 新着記事	目先の利益を最優先する「PL脳」が、日本企業を低迷させる本質的理由【動画】 - ファイナンス思考	https://diamond.jp/articles/-/305751		2022-07-04 04:50:00
ビジネス	ダイヤモンド・オンライン - 新着記事	ホテル業界、開業ラッシュで大賑わい！復活の裏で塗り替わる勢力図と「新たな覇者」 - ホテルの新・覇者	https://diamond.jp/articles/-/305764	ホテル業界、開業ラッシュで大賑わい復活の裏で塗り替わる勢力図と「新たな覇者」ホテルの新・覇者コロナ危機に伴うホテル不況であまたのホテルがつぶれ、開業計画も頓挫した。	2022-07-04 04:45:00
ビジネス	ダイヤモンド・オンライン - 新着記事	企業を成長させる「ファイナンス思考」体得の極意、“PL脳”が企業を滅ぼす！【動画】 - ファイナンス思考	https://diamond.jp/articles/-/305750	企業価値	2022-07-04 04:40:00
ビジネス	ダイヤモンド・オンライン - 新着記事	「狼狽の中で平静を保て」波乱相場に役立つ偉人の知恵、米著名投資家が厳選した13の名言 - 政策・マーケットラボ	https://diamond.jp/articles/-/305837	株式相場	2022-07-04 04:35:00
ビジネス	ダイヤモンド・オンライン - 新着記事	「140円」視野のドル円相場、円安材料盛り沢山の中で“円高リスク”を総点検 - 政策・マーケットラボ	https://diamond.jp/articles/-/305868	日米金利差	2022-07-04 04:30:00
ビジネス	ダイヤモンド・オンライン - 新着記事	「お金の話が苦手」な営業マンに、話術よりも必要な「たった1つ」のこと - 和田裕美のステップアップ仕事論	https://diamond.jp/articles/-/305053	和田裕美	2022-07-04 04:25:00
ビジネス	ダイヤモンド・オンライン - 新着記事	年収が高い不動産会社ランキング2021最新版【トップ5】三井不、三菱地所に大差の1位は？ - ニッポンなんでもランキング！	https://diamond.jp/articles/-/305643	三菱地所	2022-07-04 04:20:00
ビジネス	ダイヤモンド・オンライン - 新着記事	年収が高い不動産会社ランキング2021最新版【全国137社完全版】 - ニッポンなんでもランキング！	https://diamond.jp/articles/-/305642	不動産会社	2022-07-04 04:20:00
ビジネス	ダイヤモンド・オンライン - 新着記事	日経平均「2026年に4万円超え」シナリオも、人気ストラテジストが描く5年後の日本株 - 今週の週刊ダイヤモンド　ここが見どころ	https://diamond.jp/articles/-/305804	日経平均「年に万円超え」シナリオも、人気ストラテジストが描く年後の日本株今週の週刊ダイヤモンドここが見どころ『週刊ダイヤモンド』月日号の第特集は「業種社年後の業界地図」です。	2022-07-04 04:15:00
ビジネス	ダイヤモンド・オンライン - 新着記事	参院選、若者が「白紙投票」でもいいから選挙に行くべき理由 - 重要ニュース解説「今を読む」	https://diamond.jp/articles/-/305624	重要	2022-07-04 04:10:00
ビジネス	ダイヤモンド・オンライン - 新着記事	山形新幹線は新幹線ではない？早期開業を可能にした「法律の抜け穴」とは - News&Analysis	https://diamond.jp/articles/-/305834	newsampampanalysis	2022-07-04 04:05:00
ビジネス	東洋経済オンライン	東海道新幹線｢3つのカイゼン｣で目指す将来の姿 速度向上だけでなく安全･安定･サービスも | 新幹線 | 東洋経済オンライン	https://toyokeizai.net/articles/-/600799?utm_source=rss&utm_medium=http&utm_campaign=link_back	東洋経済オンライン	2022-07-04 04:30:00