How to colorize black & white photos with just 100 lines of neural network code
By Emil Wallner Earlier this year, Amir Avni used neural networks to troll the subreddit/r/Colorization — a community where people colorize historical black and white images manually using Photoshop. They were astonished with Amir’s deep learning bot. What could take up to a month of manual labour could now be done in just a few seconds. I was fascinated by Amir’s neural network, so I reproduced i
これから強化学習を勉強する人のための「強化学習アルゴリズム・マップ」と、実装例まとめ - Qiita
Deleted articles cannot be recovered. Draft of this article would be also deleted. Are you sure you want to delete this article? ※2018年06月23日追記 PyTorchを使用した最新版の内容を次の書籍にまとめました。 つくりながら学ぶ! 深層強化学習 ~PyTorchによる実践プログラミング~ 18年6月28日発売 これから強化学習を勉強したい人に向けて、「どんなアルゴリズムがあるのか」、「どの順番で勉強すれば良いのか」を示した強化学習アルゴリズムの「学習マップ」を作成しました。 さらに、各手法を実際にどう実装すれば良いのかを、簡単な例題を対象に実装しました。 本記事では、ひとつずつ解説します。 オレンジ枠の手法は、実装例を紹介します。 ※今回マップを作るに
mixup: Beyond Empirical Risk Minimization
Large deep neural networks are powerful, but exhibit undesirable behaviors such as memorization and sensitivity to adversarial examples. In this work, we propose mixup, a simple learning principle to alleviate these issues. In essence, mixup trains a neural network on convex combinations of pairs of examples and their labels. By doing so, mixup regularizes the neural network to favor simple linear
A multi-branch convolutional neural network for detecting double JPEG compression
Detection of double JPEG compression is important to forensics analysis. A few methods were proposed based on convolutional neural networks (CNNs). These methods only accept inputs from pre-processed data, such as histogram features and/or decompressed images. In this paper, we present a CNN solution by using raw DCT (discrete cosine transformation) coefficients from JPEG images as input. Consider
Announcing New AWS Deep Learning AMI for Amazon EC2 P3 Instances
The new AMIs are pre-installed and configured with CUDA 9 for both Ubuntu and Amazon Linux, along with the other GPU drivers to take advantage of the speed of Volta on P3, including cuDNN 7.0, NCCL 2.0.5 and the NVIDIA Driver 384.81. In addition to the latest drivers, the AMI also includes popular frameworks which have been optimized for P3 and Volta including Apache MXNet v0.12 RC1, Caffe2 v0.8.1
Word embeddings in 2017: Trends and future directions
Word embeddings in 2017: Trends and future directions Word embeddings are an integral part of current NLP models, but approaches that supersede the original word2vec have not been proposed. This post focuses on the deficiencies of word embeddings and how recent approaches have tried to resolve them. This post discusses the deficiencies of word embeddings and how recent approaches have tried to res
2018 Intern Results at Preferred Networks (Part 1) - Preferred Networks Research & Development
This summer, Preferred Networks accepted a record number of interns in Tokyo from all over the world. They tackled challenging tasks around artificial intelligence together with PFN mentors. We appreciate their passion, focus, and designation to the internship. In this post, we would like to share some of their great jobs (more to come). Learning to flip objects using guided policy search Takeshi
Hey Siri: An On-device DNN-powered Voice Trigger for Apple’s Personal Assistant
Hey Siri: An On-device DNN-powered Voice Trigger for Apple’s Personal Assistant The "Hey Siri" feature allows users to invoke Siri hands-free. A very small speech recognizer runs all the time and listens for just those two words. When it detects "Hey Siri", the rest of Siri parses the following speech as a command or query. The "Hey Siri" detector uses a Deep Neural Network (DNN) to convert the ac
Lecture 15 | Efficient Methods and Hardware for Deep Learning
In Lecture 15, guest lecturer Song Han discusses algorithms and specialized hardware that can be used to accelerate training and inference of deep learning workloads. We discuss pruning, weight sharing, quantization, and other techniques for accelerating inference, as well as parallelization, mixed precision, and other techniques for accelerating training. We discuss specialized hardware for deep
ニューラルネットワークのパラメータ設定方法(scikit-learnのMLPClassifier)
あらすじ ニューラルネットワークを作成する際に、層の数、ニューロンの数、活性化関数の種類等考えるべきパラメータは非常に多くあります。 そこで、これらのパラメータがどのようにモデルや学習に影響を与えるかということをscikit-learnの MLPClassifier を使って解説したいと思います。 MLPClassifierを使うと、非常に簡単にニューラルネットワークを使うことができます。 今回はそれぞれのパラメータの意味と使い方及び各種メソッドの解説していきたいと思います。 ちなみに、scikit-learnの推定器の選び方に関しては、scikit-learn（機械学習）の推定器：Estimatorの選び方 をご参照下さい。 1. hidden_layer_sizes｜ 層の数と、ニューロンの数を指定 default : (100,) 隠れ層の層の数と、ニューロンの数をタプルで指定します
Map-based Multi-Policy Reinforcement Learning: Enhancing Adaptability of Robots by Deep Reinforcement Learning
In order for robots to perform mission-critical tasks, it is essential that they are able to quickly adapt to changes in their environment as well as to injuries and or other bodily changes. Deep reinforcement learning has been shown to be successful in training robot control policies for operation in complex environments. However, existing methods typically employ only a single policy. This can l
TechCrunch | Startup and Technology News
Unlike Light’s older phones, the Light III sports a larger OLED display and an NFC chip to make way for future payment tools, as well as a camera.
リリース、障害情報などのサービスのお知らせ
最新の人気エントリーの配信
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
GitHub - astorfi/CNTK-World: :earth_americas: Simple and ready-to-use deep learning examples for the Microsoft Cognitive Toolkit (CNTK)
Reddit - Dive into anything
GitHub - onnx/onnx: Open standard for machine learning interoperability
Yu Yamaguchi / 山口 祐 on Twitter: "AlphaGo Zeroが72時間で人間超えてやべーってなってますけど、真にやべーのはこの学習を3日で終わらせるGoogleのマシンパワーですから。 私のPC（GTX 1080 x2）だと、同じ学習を完了するのに62年程度かかります。Master超えバージョンなら735年です。"
インテル、AI向け新プロセッサ「Nervana NNP」を発表 | スラド ハードウェア
AlphaGo Zero: Starting from scratch
AWS Deep Learning AMI Now Supports PyTorch, Keras 2 and Latest Deep Learning Frameworks
Nao Tokui / 徳井直生 on Twitter: "Googleの担当するJeff Dean氏のトーク. 以前は50万行以上あったGoogle翻訳のソースコードが、Deep Learning導入によって約500行にまで削減でできたらしい. https://t.co/I9xstDsPCL"
