2018年2月10日、恵比寿ガーデンプレイスザ・ガーデンホールにて、「Cookpad TechConf 2018」が開催されました。Cookpadのエンジニアやデザイナーがどのようにサービス開発に取り組んでいるのか、またその過程で得た技術的知見について公開します。続いて登場したのは菊田遥平氏。「Solve "unsolved" image recognition problems in service applications」と題して、人間以上の精度を出すとされていた機械学習による画像認識の理想と現実、そしてそれにまつわる「未解決」な問題の解き方を語ります。 画像分析にはどんな困難があるか 菊田遥平氏（以下、菊田）：Thanks Polly. みなさんだいぶお疲れですか？ 発表はあと2人なのでもうちょっとがんばってください。私が一番おもしろい話をする。 （会場笑） ……ようにがんばるので
CEATEC JAPAN 2009のOKIセミコンダクタのブース内で、手を振り回している人がいたので何をしているのかと思って近づいた見たところ、空中で手を振り回すと動きを認識してモニター上で照明操作ができる画像認識LSIのデモだったようです。が、単純にジェスチャーを解釈するわけではなく、モニターに向かって手を振るとスイッチがモニター内に仮想的に出現、その仮想スイッチを操作するという一風変わったシロモノでした。 詳細は以下から。 CEATEC2009 | トピックス | OKIセミコンダクタ http://www.okisemi.com/jp/topics/ceatec2009.htm OKIのブース ジェスチャーで照明を操作します、という説明 これだけ見ると何をしようとしているのかが皆目検討不能 あっちにあるのが画像認識LSI で、ジェスチャーをするとモニター内にスイッチが登場 実際にはこ
Bringing AI to Excel—4 new features announced today at Ignite | Microsoft 365 Blog
Excel’s power comes from its simplicity. At its core, Excel is three things: cells of data laid out in rows and columns, a powerful calculation engine, and a set of tools for working with the data. The result is an incredibly flexible app that hundreds of millions of people use daily in a wide variety of jobs and industries around the world. Today, we’re pleased to announce four new artificial int
Apple acquires Turi, a machine learning company | TechCrunch
Word just started going around the rumormill that Apple has acquired Turi, a company that describes itself as a “machine learning platform for developers and data scientists.” We reached out to Apple for confirmation, and sure enough — we got the standard reply they give when they’re confirming an acquisition but not saying much else: “Apple buys smaller technology companies from time to time, and
The AI superstars at Google, Facebook, Apple—they all studied under this guy
Mr. Robot Geoffrey Hinton spent 30 years hammering away at an idea most other scientists dismissed as nonsense. Then, one day in 2012, he was proven right. Canada’s most influential thinker in the field of artificial intelligence is far too classy to say I told you so By Katrina Onstad | Photograph by Daniel Ehrenworth | January 29, 2018 For more than 30 years, Geoffrey Hinton hovered at the edges
3つの要点 ✔️ 新しいadaptive gradient clipping法を用いたバッチ正規化の代替 ✔️ 正規化なしアーキテクチャNFNetsがSOTAを達成 ✔️ バッチ正規化を用いたモデルよりも、優れた学習速度と伝達学習能力を持つ High-Performance Large-Scale Image Recognition Without Normalization written by Andrew Brock, Soham De, Samuel L. Smith, Karen Simonyan (Submitted on 11 Feb 2021) Comments: Accepted to arXiv. Subjects: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Ma
Jeff Dean on Large-Scale Deep Learning at Google - High Scalability -
If you can’t understand what’s in information then it’s going to be very difficult to organize it. This quote is from Jeff Dean, currently a Wizard, er, Fellow in Google’s Systems Infrastructure Group. It’s taken from his recent talk: Large-Scale Deep Learning for Intelligent Computer Systems. Since AlphaGo vs Lee Se-dol, the modern version of John Henry’s fatal race against a steam hammer, has ca
The Python Extension for VS Code Insiders is excited to announce the new preview for Native Notebooks! Native Notebooks are VS Code’s newest implementation of notebooks and the Python Extension is leveraging the Native Notebooks API to revamp the data science experience! Users can now benefit from the new functionalities below: The same extensions that users rely on when editing source code or mar
We are an image recognition based mobile marketing company. Our Snap.Send.Get™ solution converts any image into a 100% opt-in interactive mobile ad... » READ MORE
Working as a core maintainer for PyTorch Lightning, I've grown a strong appreciation for the value of tests in software development. As I've been spinning up a new project at work, I've been spending a fair amount of time thinking about how we should test machine learning systems. A couple weeks ago, one of my coworkers sent me a fascinating paper on the topic which inspired me to dig in, collect
AutoStitch
Welcome to AutoStitch. If you have an iPhone, please check out our new iPhone version of AutoStitch below! If you're looking for the Windows demo version, you can download it using the link above, or read on to find out more about AutoStitch. Thanks for visiting! AutoStitch now brings the latest in image recognition technology to your iPhone. Stitch images in any order or arrangement, using photos
画像認識AIカオスマップ2023年最新版を公開！
企業のDXを推進する国内最大級のAIポータルメディア「AIsmiley」を運営するアイスマイリー（東京都渋谷区、代表取締役：板羽晃司）は、各業界のDX推進の支援の一環として、画像認識AI関連サービスをまとめた「画像認識カオスマップ2023」を2023年2月28日に公開しました。掲載数は合計で204サービスです。 ■画像認識AIカオスマップ2023 こちらのカオスマップは画像認識AIを業界別で探せるように「製造」「自動車」「建設」「エネルギー」「防犯」「物流」「医療機関」等のカテゴリーに分け、合計204サービスをマッピングしております。作成にあたり参考にしたサービスURL、画像認識AIベンダーを記載した一覧表（Excel）は、カオスマップ資料請求後に画像認識AIの導入を検討している企業ご担当者様に無償でご案内いたします。 ■画像認識AIとは？ 画像認識AIとは、AI・人工知能が人間の目の代わ
How HBO’s Silicon Valley built “Not Hotdog” with mobile TensorFlow, Keras & React Native | HackerNoon
How HBO’s Silicon Valley built “Not Hotdog” with mobile TensorFlow, Keras & React Native Too Long; Didn't ReadThe <a href="http://www.hbo.com/silicon-valley">HBO show <em>Silicon Valley</em></a> released a real AI app that identifies hotdogs — and not hotdogs — like the one shown on season 4’s 4th episode (the app is <a href="https://www.seefoodtechnologies.com/nothotdog/">now available on Android
The Blog of Scott Aaronson If you take nothing else from this blog: quantum computers won't solve hard problems instantly by just trying all solutions in parallel. Wrap-Up (June 5): This will be my final update on this post (really!!), since the discussion seems to have reached a point where not much progress is being made, and since I’d like to oblige the commenters who’ve asked me to change the
RaiMan's SikuliX
Automate what you see on a computer monitor What is it? For what is it? Get it? Use it? Get help? Contribute? Latest stable version: 2.0.5 —— Nightly builds: currently not available A new webpage is on the way -- having a new SikuliX logo ... visit the news blog - enjoy! Sikuli is God's Eye … in Huichol Indian culture: the power to see and understand things un
A personal Google, just for you
When Google was founded, there were about 300 million people using the Internet. The vast majority of them were sitting at desktop computers and looking for answers that came in the form of blue links. Today, the Internet community is closer to 3 billion people, and you’re searching for all kinds of help everywhere — from your cars and your classrooms, to your homes, to the phones in your pockets.
Credit for the title image: Liu et al. (2021) 2021 saw many exciting advances in machine learning (ML) and natural language processing (NLP). In this post, I will cover the papers and research areas that I found most inspiring. I tried to cover the papers that I was aware of but likely missed many relevant ones. Feel free to highlight them as well as ones that you found inspiring in the comments.
こんにちは、データチームの後藤です。 VASILYデータチームは2017年8月7日〜10日にかけて、広島で行われた第20回画像の認識・理解シンポジウム（以下、MIRU2017）に参加しました。本記事では、発表の様子や参加した感想をお伝えしたいと思います。 MIRU2017 MIRUはMeeting on Image Recognition and Understandingの略で、国内最大規模の画像の認識と理解技術に関する会議です。事前に選定された口頭発表と国際会議の採択論文から選ばれた招待講演を中心に、ポスター発表、デモ発表、特別講演で構成されます。 今年の投稿件数は、以下のような内訳でした。 口頭発表：23件 一般論文：206件 デモ論文：13件 招待講演：12件 企業展示：15件 VASILYのデータチームは「トリプレット損失関数の重み付けによる学習の効率化」というタイトルで一般論文
IQ Engines: Image Recognition and Visual SearchIQ Engines recognizes and labels photos, monetizing images by matching them to relevant retailers and advertisers. Our image recognition engine is fast, accurate and scalable. We provide image intelligence via an open Developer API to retailers, photo publishers and mobile app developers.
Building an AI Chip Saved Google From Building a Dozen New Data Centers
Google operates what is surely the largest computer network on Earth, a system that comprises custom-built, warehouse-sized data centers spanning 15 locations in four continents. But about six years ago, as the company embraced a new form of voice recognition on Android phones, its engineers worried that this network wasn't nearly big enough. If each of the world's Android phones used the new Goog
Deep Learning Enables You To Hide Screen When Your Boss Is Approaching - AHO Grammer
Introduction When you are working, you have browsed information that is not relevant to your work, haven’t you? I feel awkward when my boss sneaks up behind me. While I can quickly switch screens, such behavior looks suspicious, and sometimes I don’t notice him approaching. To avoid this issue without raising suspicion, I created a system that automatically recognizes when he is approaching and hi
This is a table documenting some of the best results some paper obtained in CIFAR-10 dataset. Spatially-sparse convolutional neural networks (ARXIV 2014) Cited 12 times. 93.72% Additional info: DeepCNiN(5,300) With data augmentation and Network-in-Network layers Deep Residual Learning for Image Recognition (ARXIV 2015) Cited 1 time. 93.57% Additional info: ResNet 110 layers, 1.7 million parameters
アマゾン傘下の検索技術企業 A9 が、画像認識技術を用いた iOS 用 AR アプリ " Flow Powered by Amazon " を公開しました。書籍やDVD、ゲームなどに iOS端末のカメラを向けると画像認識により自動的にAmazon.comのデータベースに問い合わせ、価格や評価といった情報をポップアップ表示します。 面白いのはバーコードだけでなく、表紙やパッケージをかなりの速度・精度で認識してくれること。バーコードで蔵書を管理したりネット上の商品データベースを引くアプリは1ジャンルをなすほど歴史があり、また画像認識といえば Google モバイルアプリの Google Goggles が有名ですが、" Flow " はシャッターを切って認識を待つ必要もなく、カメラを向ければシームレスかつ連続的に認識が可能です。 また検索結果ページに飛ぶのではなく、画面はリアルタイムに対象
February 4, 2016 by Sam Gross and Michael Wilber The post was co-authored by Sam Gross from Facebook AI Research and Michael Wilber from CornellTech. In this blog post we implement Deep Residual Networks (ResNets) and investigate ResNets from a model-selection and optimization perspective. We also discuss multi-GPU optimizations and engineering best-practices in training ResNets. We finally compar
Applying Deep Learning to Enhance Momentum Trading Strategies in Stocks
This version: December 12, 2013 Applying Deep Learning to Enhance Momentum Trading Strategies in Stocks Lawrence Takeuchi * ltakeuch@stanford.edu Yu-Ying (Albert) Lee yy.albert.lee@gmail.com Abstract We use an autoencoder composed of stacked restricted Boltzmann machines to extract features from the history of individual stock prices. Our model is able to discover an en- hanced version of the mome
Image recognition (i.e. classifying what object is shown in an image) is a core task in computer vision, as it enables various downstream applications (automatically tagging photos, assisting visually impaired people, etc.), and has become a standard task on which to benchmark machine learning (ML) algorithms. Deep learning (DL) algorithms have, over the past decade, emerged as the most competitiv
python-machine-learning-book/code at master · rasbt/python-machine-learning-book
Simply click on the ipynb/nbviewer links next to the chapter headlines to view the code examples (currently, the internal document links are only supported by the NbViewer version). Please note that these are just the code examples accompanying the book, which I uploaded for your convenience; be aware that these notebooks may not be useful without the formulae and descriptive text. Machine Learnin
If you have additions or changes, send an e-mail (remove the "nospam"). This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each authors copyright. Session 1: Recognition A
3つの要点 ✔️ 画像分類タスクに必要不可欠なData Augmentationの体系をまとめた ✔️ 基本的なData Augmentationについて手法と利点/欠点をまとめた ✔️ 基本的なDAは実装が簡単な上に絶大な効果を発揮する A survey on Image Data Augmentation for Deep Learning written by Connor Shorten, Taghi M. Khoshgoftaar (Submitted on 06 July 2019) Comments: Published by Journal of Big Data Subjects: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Image and Video Proc
Adversarial Machines
Adversarial A.Is are a common sci-fi theme: Robot VS Robot. In recent years, real adversarial examples have emerged. This experiment explores how to generate images to fool A.Is (and turn everyone into manga). Convolutional Neural NetworksAt the heart of many modern computer vision systems are Convolutional Neural Networks. On some vision tasks, CNNs have surpassed human performance. Industries su
Artificial intelligence video tools make it relatively easy to put one person’s face on another person’s body with few traces of manipulation. I tried it on myself. What could go wrong? The scene opened on a room with a red sofa, a potted plant and the kind of bland modern art you’d see on a therapist’s wall. In the room was Michelle Obama, or someone who looked exactly like her. Wearing a low-cut
Capture the flag. Hide and seek. Marco Polo. These location-based games brought hours of fun to many of us as children. Then video games came along and suddenly the only location you played in was the living room. Now this shift is coming full circle as innovative mobile games are using geo-location, image recognition and augmented reality technologies to combine the real and virtual worlds. Locat
![Why Location-Based Gaming Is The Next Killer App [OPINION]](https://cdn-ak-scissors.b.st-hatena.com/image/square/99a9d941acf4ba37b578789ed87ae21c22bcb4b4/height=288;version=1;width=512/https%3A%2F%2Fhelios-i.mashable.com%2Fimagery%2Farchives%2F037MBbM0M4ozEtLzuBrCIHZ%2Fhero-image.fill.size_1200x675.v1647020093.jpg)
New to Microsoft 365 in February—advancing security and empowering a modern workplace | Microsoft 365 Blog
This month, we released several new capabilities to help you stay ahead of threats, create a more productive workplace, and keep you in the flow of work. Here’s a look at what’s new in February. Stay ahead of threats and collaborate securely New features and services help you better manage a complex threat landscape and communicate and collaborate securely. Extend your security team’s capability w
Reading the VGG Network Paper and Implementing It From Scratch with Keras | HackerNoon
Too Long; Didn't ReadThere are hundreds of code examples for Keras. It's common to just copy-and-paste code without knowing what's really happening. In this tutorial, you will implement something very simple, but with several learning benefits: you will implement the VGG network with Keras, from scratch, by reading the VGG's* original paper. There are hundreds of code examples for Keras. It's comm
＜サービス支配論理＞ iTunesに対する一発逆転を狙ったアマゾンの音楽サービス「クラウドドライブ」が引き起こした激しい波紋！！: SNS,ソーシャルネットワーキング.jp
＜サービス支配論理＞ iTunesに対する一発逆転を狙ったアマゾンの音楽サービス「クラウドドライブ」が引き起こした激しい波紋！！ ＜炎上＞ イオングループ企業、取締役なりすましTwitterが原因で炎上 「法的措置も」 ＜スマートテレビ＆ソーシャルテレビ＞ 近未来のスマートテレビの動画 ＜スマートテレビ＞YouTubeがスマートテレビ時代に合わせて変化中！！ ＜ネット世代論＞ デートの様子をネットで中継！？ ネット世代の「リアルな関係」とは何か ＜スマートテレビ＞米国ディズニーのスポーツチャネルＥＳＰＮがiPad視聴アプリ発表！！ ＜アップス＞ スポーツＥＳＰＮのiPad視聴アップスの動画！！ ＜マイクロ取引＞米国新聞トップのＵＳＡトウディが記事へのアクセス数をボーナスに反映か？ ＜プロダクトプレースメント＞ リーバイスがフェースブックのソーシャルゲームMall World
NeurIPS 2022 参加報告 後編
はじめに プロダクトオーナー兼機械学習エンジニアの本田志温です。 弊社高橋による前回の記事「NeurIPS 2022 参加報告 前編」 に引き続き、同会議の参加報告をします。本記事では、個人的に気になった論文（計53本）をいくつかのカテゴリで分類し、カテゴリごとに研究トレンドを大づかみにできるような形で書きます。特に重要だと感じた論文は詳しめに取り上げます。 会場の様子 また、本記事に関心をお持ちになった方は以下の過去記事もお楽しみいただけるのではないかと思います。ぜひ合わせてご覧ください。 AI開発の新たなパラダイム「基盤モデル」とは NeurIPS 2021 参加報告 前編 NeurIPS 2021 参加報告 後編 深層学習の原理 深層学習は様々なタスクで高い性能を発揮することが経験的に知られていますが、「なぜうまくいくのか」という原理についてわかっていることは多くありません。そのため
PyTorch developer ecosystem expands, 1.0 stable release now available - Facebook Code
PyTorch developer ecosystem expands, 1.0 stable release now available As the PyTorch ecosystem and community continue to grow with interesting new projects and educational resources for developers, today at the NeurIPS conference we’re releasing PyTorch 1.0 stable. The latest version, which was first shared in a preview release during the PyTorch Developer Conference in October, includes capabilit
Machine learning libraries are becoming faster and more accessible with each passing year, showing no signs of slowing down. While traditionally Python has been the go-to language for machine learning, nowadays neural networks can run in any language, including JavaScript! The web ecosystem has made a lot of progress in recent times and although JavaScript and Node.js are still less performant tha
In probability theory, the Chinese restaurant process is a discrete-time stochastic process, analogous to seating customers at tables in a restaurant. Imagine a restaurant with an infinite number of circular tables, each with infinite capacity. Customer 1 sits at the first table. The next customer either sits at the same table as customer 1, or the next table. This continues, with each customer ch
functional APIでKerasを始めてみよう functional APIは，複数の出力があるモデルや有向非巡回グラフ，共有レイヤーを持ったモデルなどの複雑なモデルを定義するためのインターフェースです． ここではSequentialモデルについて既に知識があることを前提として説明します． シンプルな例から見てきましょう． 例1: 全結合ネットワーク 下記のネットワークはSequentialモデルによっても定義可能ですが， functional APIを使ったシンプルな例を見てきましょう． レイヤーのインスタンスは関数呼び出し可能で，戻り値としてテンソルを返します Modelを定義することで入力と出力のテンソルは接続されます 上記で定義したモデルはSequentialと同様に利用可能です from keras.layers import Input, Dense from kera
j次のブックマーク
k前のブックマーク
lあとで読む
eコメント一覧を開く
oページを開く
機械学習を用いた画像分類における「未解決問題」を解くためにやったこと
ジェスチャーでモニター上にスイッチが出現、動きを認識して照明操作できる画像認識LSIのデモムービー
バッチ正規化を使用しない新たなチャンピオンNFNetsが登場！画像タスクトップの性能！
Notebooks are getting revamped! - Python
SnapTell - Image recognition based mobile marketing
Effective testing for machine learning systems.
D-Wave: Truth finally starts to emerge
ML and NLP Research Highlights of 2021
MIRU2017参加報告 - ZOZO TECH BLOG
IQ Engines: Image Recognition and Visual Search
Table of results for CIFAR-10 dataset
Engadget Japanese for Mobile
Torch | Training and investigating Residual Nets
A Visual History of Interpretation for Image Recognition
ICCV 2011 papers on the web - Papers
【基本編】画像認識に使用されるData Augmentationを一挙にまとめてみた！
Here Come the Fake Videos, Too (Published 2018)
Why Location-Based Gaming Is The Next Killer App [OPINION]
10 Machine Learning Examples in JavaScript
Chinese restaurant process - Wikipedia
Functional APIのガイド - Keras Documentation