History of DEEP COGNITION - Deep Learning Studio

Last update Apr. 4^th 2018 10:22:42 AM

Creation date Mar. 25^th 2018 10:14:35 PM

• 
  PALYGAP on Apr. 4^th 2018 10:22:42 AMReport

  Development Activity

  Important (2 enhanced versions have been release recently)

• 
  PALYGAP on Apr. 3^rd 2018 9:26:56 PMReport

  Live performance monitoring

  yes

  Results browser

  yes

  View Valid/test set inference

  yes

  Confusion matrix

  no

  Visualize intermediate layer activation

  yes

  User test int

  yes

  - Cloud

  yes

  but no computing environnement"

  yes

  - Cloud Trial period

  yes

  AutoML

  yes

  Browser

  yes

  Python shell

  yes

  - Desktop

  yes (any platform with python)

• 
  PALYGAP on Apr. 3^rd 2018 9:22:04 PMReport

  Custom cost function

  yes

  Load Keras Models + Weights

  yes

  Load Caffe Models + Weights

  no

  Model checker

  yes

  Experiments Mangement

  yes

  - Exp comparison

  yes

• 
  PALYGAP on Apr. 3^rd 2018 9:20:52 PMReport

  - LSTM

  yes

  - MLP / Auto-encoders

  yes

  - RNN

  yes

  Use Keras API

  yes

  Save Keras HDFS file (.h5)

  yes

  Save Keras model .yaml file

  no

  Save Keras model file

  no

  Load Keras model .yaml file

  yes

  Load Keras HDFS file (.h5)

  yes

  Code generation

  no

  - Theano

  no

  - Tensorflow/Keras

  no

  - Caffe (v1)

  no

• 
  PALYGAP on Mar. 30^th 2018 5:31:33 AMReport

  Tune learning parameter

  yes

  DL model training

  yes

  Inference available

  yes

  Graphical editor provided

  yes

• 
  PALYGAP on Mar. 26^th 2018 9:05:35 PMReport

  Name

  DEEP COGNITION - Deep Learning Studio

• 
  PALYGAP on Mar. 25^th 2018 10:39:43 PMReport

  Web Site

  http://www.deepcognition.ai

• 
  PALYGAP on Mar. 25^th 2018 10:37:30 PMReport

  Web Site

  http://www.deepcognition.com

• 
  PALYGAP on Mar. 25^th 2018 10:14:36 PMReport

  Mode

  70

  Name

  Deep Learning Studio

  Viewer

  public

  Editor

  palygap-1q7klct2

  Version of the software

  2.0

  Underlaying Framework

  Apache MXNet

  Development Activity

  "Important

  (2 enhanced versions have been release recently)"

  unknown

  GPUs

  CUDA compatible

  DL MODEL EDITOR

  Yes

  Changing the model is possible by editing the network graph"

  DLS Better

  - CNN

  yes

  - LSTM

  Yes

  - MLP / Auto-encoders

  Yes

  - RNN

  Yes

  Pretained Networks

  yes (Resnet50, InceptionV3, SqueezeNet, VGG16, VGG19, WideResNet)

  Use Keras API

  Yes

  Save Keras HDFS file (.h5)

  Yes

  Save Keras model .yaml file

  No

  Save Keras model file

  No

  Load Keras model .yaml file

  Yes

  Load Keras HDFS file (.h5)

  Yes

  Code generation

  No

  - Theano

  No

  - Tensorflow/Keras

  No

  - Caffe (v1)

  No

  Custom cost function

  Yes

  Load Keras Models + Weights

  Yes

  Load Caffe Models + Weights

  No

  Model checker

  Yes

  Datasets Management

  yes

  Experiments Mangement

  Yes

  - Exp comparison

  Yes

  - Prev. Exp loader

  Yes

  TRAINING

  yes

  Tune learning parameter

  Yes

  Live performance monitoring

  Yes

  Results browser

  Yes

  INFERENCE

  yes

  View Valid/test set inference

  Yes

  Confusion matrix

  No

  Visualize intermediate layer activation

  Yes

  User test int

  Yes

  DEPLOIEMENT

  yes

  - Rest API

  yes

  - Cloud

  Yes

  but no computing environnement"

  Yes

  - Cloud Trial period

  Yes

  - Desktop

  Yes (any platform with python)

  AutoML

  Yes

  Browser

  Yes

  Environment management

  yes

  Python shell

  Yes

  Jupyter notebook

  yes

  Documentation

  "Minimal

  (Keras API)"

  "Minimal

  (SONY Neural Network Library)"

  ???

  Tutorials

  yes