pmigdal + deep-learning   235

Understanding Hinton’s Capsule Networks. Part I: Intuition.
Last week, Geoffrey Hinton and his team published two papers that introduced a completely new type of neural network based on so-called capsules. In addition to that, the team published an algorithm…
capsnet  deep-learning  cnn 
3 days ago by pmigdal
Using deep learning for Single Image Super Resolution
We apply three different deep learning models to reproduce stare-of-the-art results in single image super resolution.
deep-learning  upscaling  upsampling 
17 days ago by pmigdal
Experiments with a new kind of convolution – Towards Data Science – Medium
There are many things that I don’t like about convolution. The biggest of them all, most of the weights particularly in later layers are quite close to zero. This tells that most of these weights…
deep-learning  convolution 
5 weeks ago by pmigdal
Human log loss for image classification | deepsense.ai Blog
Deep learning vs human perception: creating a log loss benchmark - my post based on @kaggle cervical cancer contest
deep-learning  log-loss  human-learning 
9 weeks ago by pmigdal
Neptune - Machine Learning Lab
Neptune by @deepsense_ai - machine learning lab & experiment tracking - now gives $100 for GPU cloud computing
cloud-computing  machine-learning  deep-learning  version-control 
12 weeks ago by pmigdal
ImageNet: the data that spawned the current AI boom — Quartz
In 2006, Fei-Fei Li started ruminating on an idea. Li, a newly-minted computer science professor at University of Illinois Urbana-Champaign, saw her colleagues across academia and the AI industry hammering away at the same concept: a better algorithm would make better decisions, regardless of the data. But she realized a limitation to this approach—the best...
deep-learning  imagenet  dataset 
july 2017 by pmigdal
37 Reasons why your Neural Network is not working – Slav
The network had been training for the last 12 hours. It all looked good: the gradients were flowing and the loss was decreasing. But then came the predictions: all zeroes, all background, nothing…
deep-learning  debugging 
july 2017 by pmigdal
How to Visualize Your Recurrent Neural Network with Attention in Keras
Visualizing recurrent NNs for machine translation using "attention" in keras https://t.co/P5OsgD7zUb. Blog post coming soon! @your_datalogue
rnn  keras  deep-learning 
july 2017 by pmigdal
Neural Coreference - they, she I – Hugging Face
State-of-the-art neural coreference resolution system
nlp  pronoun  deep-learning  spacy 
july 2017 by pmigdal
Semantic Segmentation using Fully Convolutional Networks over the years
Semantic Segmentation using Fully Convolutional Networks over the years

(U-Net, DenseNet, Mask R-CNN, ...)
deep-learning  u-net  image-segmentation 
july 2017 by pmigdal
Cardiologist-Level Arrhythmia Detection With Convolutional Neural Networks
Diagnosing arrhythmias from single-lead ECG signals better than a cardiologist.
ecg  deep-learning 
july 2017 by pmigdal
[D] Why do people draw neural networks upside down? : MachineLearning
Why do people draw neural networks upside down?
Even decision trees grow from their roots downwards!

flow  deep-learning  metaphor  tree  from twitter
june 2017 by pmigdal
General Game Playing with Schema Networks
Vicarious introduces the Schema Network, a generative graphical model that can simulate the future and reason about cause and effect. We demonstrate the benefits of this kind of reasoning for game playing and show an adaptability not seen before in other agents.
deep-learning  q-learning 
june 2017 by pmigdal
Exploring LSTMs
RT @echen: Exploring LSTMs: tutorial + discovery Finally moving my blog into m…
lstm  deep-learning 
june 2017 by pmigdal
AlphaGo, in context – Andrej Karpathy – Medium
I had a chance to talk to several people about the recent AlphaGo matches with Ke Jie and others. In particular, most of the coverage was a mix of popular science + PR so the most common questions I…
go  q-learning  deep-learning 
june 2017 by pmigdal
szagoruyko/diracnets: Training Very Deep Neural Networks Without Skip-Connections
diracnets - Training Very Deep Neural Networks Without Skip-Connections
relu  deep-learning 
june 2017 by pmigdal
Fontjoy - Get font ideas with deep learning
Fontjoy helps designers choose the best font pairs using deep learning
font  deep-learning 
may 2017 by pmigdal
handong1587.github.io/2015-10-09-dl-tutorials.md at master · handong1587/handong1587.github.io
nice convnet pictures
Contribute to handong1587.github.io development by creating an account on GitHub.
convolution  deep-learning 
may 2017 by pmigdal
2nd place solution for the 2017 national datascience bowl – Julian de Wit – Freelance software/machine learning engineer.
Summary
This document describes my part of the 2nd prize solution to the Data Science Bowl 2017 hosted by Kaggle.com. I teamed up with Daniel Hammack. His part of the solution is decribed here The goal of the challenge was to predict the development of lung cancer in a patient given a set of CT images. Detailed descriptions of the challenge can be found on the Kaggle competition page and this blog post by Elias Vansteenkiste. My solution (and that of Daniel) was mainly based on nodul...
lung  cancer  3d-conv  deep-learning  health  unet 
may 2017 by pmigdal
Learning Deep Learning with Keras by Piotr Migdał
@DataScienceLA So, here is Learning Deep Learning with Keras, with your chart: I asked & hope you like it's usage :)
keras  deep-learning 
may 2017 by pmigdal
A Brief History of CNNs in Image Segmentation: From R-CNN to Mask R-CNN
At Athelas, we use Convolutional Neural Networks(CNNs) for a lot more than just classification! In this post, we’ll see how CNNs can be used, with great results, in image instance segmentation. Ever…
rcnn  deep-learning  image-detection 
april 2017 by pmigdal
Deep Learning - What I actually do | What People Think I Do / What I Really Do | Know Your Meme
Writing Learning Deep Learning: spent one hour tracking one meme's origin (), another - creating its #Keras variant.
deep-learning  humour  meme  theano 
april 2017 by pmigdal
« earlier      
per page:    204080120160

related tags

3d  3d-conv  academia  acoustics  ai  ami  analogies  analogy  architecture  arrhythmia  art  atrial-fibrillation  autoencoder  aws  backpropagation  basic-income  batch-norm  bayesian  beauty  benchmark  binary  biology  book  bot  caffe  cancer  capsnet  career  cars  cats  cell  cervix  char2char  cheatsheet  chemistry  cloud-computing  cnn  cognitive  coloring  compression  convnet  convolution  course  cpu  curiosity  data  data-art  data-augmentation  dataset  debugging  deep-dream  deep-learning  deepdream  deepsense  demography  differentation  doom  drawing  dream  ecg  economy  ecs  expression  eye  face  facial-recognition  flow  fluid  font  forecasting  fourier  gamedev  gan  github  go  gpu  gradient-descent  grammar  graph  gui  health  history  hn  human-learning  humour  illusion  image-detection  image-generation  image-processing  image-recognition  image-segmentation  imagenet  install  interactive  interpolation  ipython-notebook  javascript  journal  jupyter  kaggle  keras  lasagne  log-loss  lstm  lua  lung  machine-learning  machinelearning  mario  matrix  medicine  meme  meta-learning  metaphor  metropolis  mushroom  music  network-visualization  neural-network  neural-networks  neural-style  news  nlp  nsfw  nude  nudity  openscience  optimization  painting  papers  performance  philosophy  photography  photos  physics  pokemon  porn  programming  progress  pronoun  prototypical  publications  python  pytorch  q-learning  quantum  rcnn  reddit  reinforcement-learning  relu  repo  resources  retina  rnn  saliency  science  sdg  segmentation  selfie  sentiment-analysis  seq2seq  society  sociology  sound  spacy  spectral  starcraft  starcraft2  tensor  tensorboard  tensorflow  test  theano  torch  translation  tree  trends  turing  tutorial  typography  u-net  uber  unet  unsupervised  upsampling  upscaling  version-control  video  vision  visualisation  visualization  webgl  whale  whales  word2vec  word2wec  xgboost 

Copy this bookmark:



description:


tags: