GitHub - qiuqiangkong/audioset_tagging_cnn
@PINTO03091
"Large-Scale Pretrained Audio Neural Networks for Audio Pattern Recognition"
t.co/mM2YyQXBnJ
レコメンデーション領域における横断データ活用の取り組み事例紹介 #機械学習 - Yahoo! JAPAN Tech Blog
@icoxfog417
推薦にユーザーの検索履歴を使用した事例。検索クエリを検索クエリの情報で事前学習した単語分散表現でベクトル化し平均ベクトルを使用するのが基本だが、推薦アイテムに沿ったクエリのみにフィルタする工夫を行っている。
t.co/jmiuBvRPXC
Dolt Use Cases | DoltHub Blog
@xeraa
"Dolt is Git for data": t.co/U6miWI858x
"the only database with branches" because didn't we all want to resolve some merge conflicts in our datastore as well 😉
on a more serious note: didn't CouchDB do something similar 10+ years ago (just without calling it git)?
Rasa Livecoding: Dialect bot (Exporting trained models from Jupyter, Part 2...
@rctatman
Speeeaking of livecoding, my next stream will be tomorrow morning at 9:00 Pacific! (Note that we did daylight savings a couple weeks ago so might have shifted relative to you.)
We'll be (hopefully) solving The Mystery Of the Mismatched Labels. 🕵️☕️🖥️
t.co/jQNe5w3l3s
Introducing the Model Garden for TensorFlow 2 — The TensorFlow Blog
@fchollet
The TensorFlow Model Garden has been redesigned for TF 2.0. It has become a lot more usable and more performant! Check it out. t.co/a40JSJp0Z8 t.co/wo86C1qWs3
Kaggleで戦いたい人のためのpandas実戦入門 - ML_BearのKaggleな日常
@MLBear2
「Kaggleでテーブルコンペやるならこれくらい知ってたら便利かな」と思っているpandasの機能をまとめてみました。(もちろん普通のデータ分析の仕事にも使えるはず)
僕も半年前まではpandas苦手マンだったので、同じように苦手意識持っている人のお役に立てれば嬉しいです。
t.co/wDmpx4YvLT
[2003.13845] AvatarMe: Realistically Renderable 3D Facial Reconstruction "i...
@ak92501
AvatarMe: Realistically Renderable 3D Facial Reconstruction “in-the-wild”
pdf: t.co/TwNRd3XpUP
abs: t.co/rwKs1q7F2E t.co/cySrt0Cgwm
[2003.13845] AvatarMe: Realistically Renderable 3D Facial Reconstruction "i...
@yshhrknmr
一枚の顔画像からのレンダリング可能な3Dモデルの構築。僕もやりたいけど高品質キャプチャデータがないと手も足も出ない…。
AvatarMe: Realistically Renderable 3D Facial Reconstruction "in-the-wild", CVPR2020
t.co/LXcUGp968r
t.co/h78eUWyu6h t.co/XSKOmLYwAv
[2003.14323] How Useful is Self-Supervised Pretraining for Visual Tasks?con...
@evolvingstuff
How Useful is Self-Supervised Pretraining for Visual Tasks?
"the greatest benefits of pretraining are currently in low data regimes, and utility approaches zero before performance plateaus on the task from additional labels."
t.co/rFJmiPimri t.co/CqVjdrBEak
2019 ACM Prize
@DeepMind
Congratulations to David Silver, recipient of this year’s @TheOfficialACM prize in Computing! David is being recognised for his role developing #AlphaGo and #AlphaZero and his contribution to the field of deep reinforcement learning. Read more here: t.co/vVlBQmo9Cz
2019 ACM Prize
@Infosys
Congratulations to David Silver for his accomplishments in the field of computer game-playing and leading #research in deep reinforcement learning. We are proud to join @TheOfficialACM in recognizing young computing professionals t.co/n7qe7EiVYV
2019 ACM Prize
@TheOfficialACM
David Silver, a Professor at University College London (@ucl) and a Principal Research Scientist at @DeepMind, will receive the 2019 ACM Prize in Computing for breakthrough advances in computer game-playing: t.co/GklniSXd0r #ACMPrize t.co/2hZOUrLeLJ
2019 ACM Prize
@ACM_CEO
Congratulations to David Silver, a Professor at @ucl and Principal Research Scientist at @DeepMind on receiving the 2019 ACM Prize in Computing for this work on computer game-playing: t.co/IPX08mCnkz
2019 ACM Prize
@souzatharsis
ACM Prize in Computing goes to University College London Professor and Alpha Go developer Prof. David Silver!
#deeplearning #datascience #machinelearning #neuralnetworks t.co/eHhUWTdB8z
2019 ACM Prize
@AnalyticaGlobal
Congratulations to David Silver for his accomplishments in the field of computer game-playing and leading #research in deep reinforcement learning. We are proud to join @TheOfficialACM in recognizing young computing professionals t.co/2jSeYPJjYl
2019 ACM Prize
@TheOfficialACM
In developing #AlphaGo, which defeated Go world champion Lee Sedol in 2016, Silver and his team at @DeepMind produced a breakthrough result that astonished the scientific world: t.co/GklniSXd0r
#ACMPrize t.co/nkTCPj7ooN
【機械学習】教師あり学習と教師なし学習の違い - YouTube
@GreenGreenMidor
ヨビノリ先生（@Yobinori ）がめちゃくちゃ分かりやすく教師あり学習、教師なし学習を解説してる！！！！！
「機械学習って何？？」
「AIってどんな事ができるの？？」
って人はこちらを確認すれば、かなりイメージが掴めると思います♪
t.co/fgOdyWzRds
[2003.14415] State-of-Art-Reviewing: A Radical Proposal to Improve Scientif...
@hardmaru
A metaphor of what state-of-the-art should really be about.
t.co/KSMK43qxBk t.co/A6PsQvAKoF
[2003.14415] State-of-Art-Reviewing: A Radical Proposal to Improve Scientif...
@rasbt
Peer review forms the backbone of modern scientific manuscript eval. [...] does this protocol remain fit for purpose in 2020? [...] we answer this question in the negative (strong reject, high confidence) and propose instead State-Of-the-Art Review (SOAR) t.co/dQrvvrDiue
[2003.14415] State-of-Art-Reviewing: A Radical Proposal to Improve Scientif...
@andrey_kurenkov
A very intriguing proposal for better reviewing 🤔
Well time to release this on April 1st! 😅 Kudos @SamuelAlbanie et al, good stuff.
t.co/YnbRQY1rhi t.co/3VHgwW7Vhs
[2003.14415] State-of-Art-Reviewing: A Radical Proposal to Improve Scientif...
@bindureddy
A completely automated way to review scientific papers!!! No more hand wringing about rejections. The all-powerful AI will review your paper😎
State of the Art Reviewing: A Radical Proposal to Improve Scientific Publication - t.co/mxK7kvfEQv
[2003.14415] State-of-Art-Reviewing: A Radical Proposal to Improve Scientif...
@dennybritz
A fun read while you’re stuck inside:
State-of-Art-Reviewing: A Radical Proposal to Improve Scientific Publication: t.co/hTalBMXKe2
“Even when the axes are rotated slightly, it remains difficult to preserve an upwards trend.” t.co/gfmzLu0Tt0
[2003.14415] State-of-Art-Reviewing: A Radical Proposal to Improve Scientif...
@ProfessorMunchy
Do you think peer-review is a little antiquated? Think we have enough AI to replace it already? Well buckle up bucko... t.co/FGyWCW9xan t.co/DbxCfmO2Zx
[1312.5602] Playing Atari with Deep Reinforcement Learningcontact arXivarXi...
@jaguring1
今回の手法は、アタリの57種類のゲーム全てに対し汎用的に適用できる手法という意味で、一つの節目だと感じた。アタリのゲームを使った汎用AI研究は、まだいくつか興味深い方向性が残されているのはもちろんだが、2013年の論文からの歴史をここら辺でいったん振り返りたくなる
t.co/J9raYDhQaj
Computer Vision Group - Kerl
@sonicair
次点でDVOかなぁ
測光誤差最小化を最小二乗法に落とし込んで解いてるだけなので中身はそんなに難しくない
著者の修士論文とぼくのブログ読めばおそらく再現実装もできるはず
t.co/iXX4lLIBSb
t.co/f4ebmkS7Cz
下町データサイエンティスト 新卒2年目が終わる - 下町データサイエンティストの日常
@nino_pira
新卒2年目振り返りブログ書きました
3年目もよろしくお願いします！！
t.co/I6vBm1IWKm
ALFRED -- A Benchmark for Interpreting Grounded Instructions for Everyday T...
@_jessethomason_
Our ALFRED leaderboard is now live!
t.co/uf51glIbNn
CVPR 2020 paper: t.co/Y3SbWsnSrq
Leaderboard: t.co/YMijxNf0pl
ECCV Workshop challenge: t.co/Je3Yv0GubJ
This benchmark enables inferring agent actions for household tasks from natural language. t.co/pYx2g0SftA
Sample-Efficient Deep Reinforcement Learning for Continuous Control
@shaneguML
It was an honor to have David on my PhD committee @CambridgeMLG. He provided precise feedback on my thesis and papers. And thank him for always asking (hard) questions when I gave talks at conferences :) Curious, respectful, and rigorous. Congratulations! t.co/8uVOHXKRk8 t.co/esNLPtfXY3 t.co/dJWecSjtEW
AI Weirdness • April Fool’s pranks written by neural network
@JanelleCShane
I tried this with a different neural net a while ago and arrived at similarly inscrutable pranks.
At least nobody will be expecting these
t.co/XWrdrIQQjT t.co/sW8ZCVBrP4
Google AI Blog: Improving Audio Quality in Duo with WaveNetEQ
@EmilStenstrom
Google now has a machine learning model that keeps talking for you (120 ms forward in time) when you’re on a video/audio call with packet loss. Sounds much better than the robotic glitches we’re used to from video conferencing. t.co/SEeonYD8wc
GitHub - ryansmcgee/seirsplus: Models of SEIRS epidemic dynamics with exten...
@gcosma1
Models of SEIRS epidemic dynamics with extensions, including network-structured populations, testing, contact tracing, and social distancing (w/ Py code). #DataScience #MachineLearning #AI #DeepLearning #Statistics #epidemiology #github
t.co/sB5WtaPoOV t.co/AOuu1NNEBW
Weights & Biases - Model explorations and hyperparameter search with W&B an...
@weights_biases
Robert Porsch and his team at Apoidea published a great end-to-end guide on training your models on a #Kubernetes cluster, and tracking them with W&B.
t.co/14CXagduZ5
[2003.13913] Flows for simultaneous manifold learning and density estimatio...
@hillbig
MFMFは低次元多様体上しかサポートがないデータに対して、多様体の推定は低次元潜在変数にノイズを加えた自己符号化器で行い、多様体上の密度推定はノイズ無のフローベースモデルで行う。低次元多様体に対して確率密度が定義できない問題に対する興味深いアプローチ t.co/wt9H1MmdKV
[2003.13913] Flows for simultaneous manifold learning and density estimatio...
@hillbig
MFMFs are new generative models for data with a low-dimensional manifold, which can provide manifold and density on the manifold. Manifold learning is achieved by AE training, while density estimation is performed by flow-based models on the manifold. t.co/wt9H1MmdKV
[2003.13913] Flows for simultaneous manifold learning and density estimatio...
@KyleCranmer
Happy to announce our most recent work "Flows for simultaneous manifold learning and density estimation" led by my friend and colleague Johann Brehmer
Paper: t.co/Ik80zDE1Xf
Code: t.co/W7j8D7LO9A t.co/5eGeSiXyey
GitHub - bayesiains/nflows: Normalizing flows in PyTorch
@driainmurray
The Neural Flows (nflows) pytorch library t.co/ivtuCbDp8q has easily-composable layers for creating many different flows (& VAEs using them), all with tests. Currently being tidied up in this new repo by its authors in Edinburgh, & as part of exciting work at @mackelab t.co/m28xbjmS7I
神経計算研究分野 | 東京大学 定量生命科学研究所
@AkiFunamizu
3月から東京大学定量生命科学研究所で講師として自分のラボを始めています。脳と機械学習をつなぐ研究を目指します。
t.co/R7uYm3idlx
Information Geometric Optimizationに対するサンプルの再利用 | AI tech studio
@cyberagent_ai
■ Research blog 公開！■
Information Geometric Optimization (IGO) という確率的最適化の枠組みと、そのIGOにサンプルの再利用の工夫を導入することで性能改善を行った論文について、#AILab 研究員の野村が解説しています！
ぜひご覧ください✏️
t.co/dMPFXKM40U
三次元点群を取り扱うニューラルネットワークのサーベイ Ver. 2 / Point Cloud Deep Learning Survey Ver. 2...
@losnuevetoros
東北大 橋本・鏡研で博士を取られた @n_chiba_ さんが、4/1からオムロンサイニックエックス (OSX) の仲間になりました！メインは早大の尾形研で、OSXにはProject Researcherとして週2日の頻度で来てくれます。
以下は千葉さんの去年のサーベイ資料です。
t.co/outL2zGNvS
Training Faster With Large Datasets using Scale and PyTorch
@keigohtr
PyTorch x Scale AIで巨大なデータセットを使った学習について。色々とTipsが詰まってる。クラウドストレージへのデータIOはメモリ効率を考えるとpythonのmultithreadingやmultiprocessingよりもasyncioの方が良いというのは知らなかった。良い。 #mlopsslack t.co/1gJUKt7fK7
BigQuery への地理空間データの読み込みを FME でより簡単に | Google Cloud Blog
@googlecloud_jp
#GoogleCloud は FME メーカー Safe Software 社とパートナーシップを締結。#BigQuery をスケーラブルな空間バックエンドとして使用できます。t.co/BAY4C7ac0O #gcpja
Understanding Deep Neural Networks: From Generalization to Interpretability...
@Equiv_Exchange
リンク先の動画、機械学習のモデルの解釈可能性について議論しているので聴いているけど、「＜スペクトル・グラフ畳込みネットワーク(GCNN)は転移学習性がない＞と言われているが、それは間違いだ」という話、＜＞をそもそも知らなかったので、感動が少ない(辛) t.co/8NCJjjqTZr
AI Academy Bootcamp | Python・機械学習・AIを実践的に学べる短期ブートキャンプ
@tankazunori0914
BootCamp Pythonプログラミング入門コースを受講されたプログラミング未経験者の方が第4回目の講義を終えて、約1ヶ月間でTwitterの自動いいねツールを自力で作れるまで成長して嬉しい😊👏
1ヶ月動画＋チャットでの質問し放題のプランは3万円なので是非受講検討お願いします！
t.co/qkgNq7RIqA
[DL輪読会]NeRF: Representing Scenes as Neural Radiance Fields for View S…
@DL_Hacks
空間の座標と視線方向を入力すると色と密度を出力するような場 (NeRF) をNNで表現。古典的なレンダリング手法との組み合わせにより任意視点の画像が生成でき、複雑なシーンでも写実的な画像が得られる。結果が衝撃的。
t.co/2mJQBMHlRZ
COVID19 Global Forecasting (Week 2) | Kaggle
@osciiart
my prediction of kaggle covid-19 week 2. I hope the curve of Italy will converge like this.
t.co/i5JoBz2VLM t.co/i5u1AzhmkF
MPRG : 機械知覚&ロボティクスグループ／中部大学
@MPRG_Chubu
2020年度の研究室メンバーを更新しました．
今年は，博士課程進学者7名(内2名 社会人ドクター)
修士課程進学者11名
情報工学科ゼミ生10名が配属され総勢64名になります．
t.co/JJCSNBvCq9
Waypoint - The official Waymo blog: Using automated data augmentation to ad...
@maxjaderberg
Improving perception models @Waymo by building on our Population Based Training style techniques to automatically tune data augmentation strategies t.co/uL1mY4NKCq t.co/oXhjAPj1zB
Waypoint - The official Waymo blog: Using automated data augmentation to ad...
@Waymo
Our newest research in collaboration with our @googleAI colleagues will allow us to train better machine learning models with less data and improve perception tasks for the Waymo Driver. t.co/vUdo5wbcN4 t.co/WvdkbhdOLs
David Mezzetti | Kaggle
@kaggle
Kaggler @DavidMezzetti has published a series of helpful #COVID19 notebooks this week. Check them out: t.co/IOKgZQQf6u t.co/VtfBXtSb5R
AI for COVID-19: An online virtual care approach
@xamat
Slides for my presentation yesterday at the AI for COVID virtual conference @StanfordHAI here: t.co/9e46NmxeFn - Full video of the conference including talk and panel discussion: t.co/XZeb98H6Lx