19
{ Deep Learning And Its Applications: Computer Vision Adam Gibson { deeplearning4j.org // skymind.io // zipfian academy

Deep Learning and its Applications - Computer Vision

Embed Size (px)

DESCRIPTION

Deep Learning and its Applications - Computer Vision Zipfian Academy Meetup Deep-learning is useful in detecting anomalies like fraud, spam and money laundering; identifying similarities to augment search and text analytics; predicting customer lifetime value and churn; recognizing faces and voices. The framework's neural nets include restricted Boltzmann machines, deep-belief networks, deep autoencoders, convolutional nets and recursive neural tensor networks.

Citation preview

Page 1: Deep Learning and its Applications - Computer Vision

{Deep Learning

And Its Applications: Computer Vision

Adam Gibson{ deeplearning4j.org // skymind.io // zipfian academy

Page 2: Deep Learning and its Applications - Computer Vision

• Object Recognition• Image Categorization• Scene Parsing• Face Recognition

Computer Vision: A Primer

Page 3: Deep Learning and its Applications - Computer Vision

• OpenCV • SIFT• Filters/Edge Detection• Feature Extraction

What’s currently done?

Page 4: Deep Learning and its Applications - Computer Vision

• Representation Learning • More precise than hand-done

features• Non-linearities and higher-

order trends• Pretrain and Hessian Free

This is manual!

Page 5: Deep Learning and its Applications - Computer Vision

• Representation Learning• Position Invariance with

convolutions• Semantic Hashing

Deep Learning and Images

Page 6: Deep Learning and its Applications - Computer Vision

• Normal pixels – 0-255 – normalization

• Sparse – binarization (depending on pixel presence)

Different kinds of images

Page 7: Deep Learning and its Applications - Computer Vision

• Faces = a collection of images.• With persistent patterns of pixels.• Pixel patterns = features.• Nets learn to identify features in data, to

classify faces as faces and label them: John or Sarah.

• Nets train by reconstructing faces from features many times.

• Measuring their work against a benchmark.

Facial recognition

Page 8: Deep Learning and its Applications - Computer Vision

DL4J’s Facial Reconstructions

Page 9: Deep Learning and its Applications - Computer Vision

• Slices of a feature space (Max pooling)• Learns different portions for easily

scalable and robust feature engineering.

Position Invariance - Convolutions

Page 10: Deep Learning and its Applications - Computer Vision

Visual Example - Convolutions

Page 11: Deep Learning and its Applications - Computer Vision

Pen Strokes

Page 12: Deep Learning and its Applications - Computer Vision

• Facebook uses facial recognition to make itself stickier and know more about us.

• Government agencies use it to secure national borders.

• Video game makers use it to construct more realistic worlds.

• Stores use it to identify customers and track behavior.

What are faces for?

Page 13: Deep Learning and its Applications - Computer Vision

• 2 layers of neuron-like nodes.• The 1st is the visible, or input, layer• The 2nd is “hidden.” It identifies features in

input• Symmetrically connected.• “Restricted” = no visible-visible or hidden-

hidden ties• All connections happen between layers.

Restricted Boltzmann Machines (RBMs)

Page 14: Deep Learning and its Applications - Computer Vision

• A stack of RBMs.• Each RBM’s hidden layer Next RBM’s

visible/input layer. • DBNs learn more & more complex features• Example:

• 1) Pixels = input; • 2) H1 learns an edge or line; • 3) H2 learns a corner or set of lines; • 4) H3 learns two groups of lines forming an

object -- a face!• Final layer classifies feature groups: sunset,

elephant, flower, John, Sarah.

Deep-Belief Net (DBN)

Page 15: Deep Learning and its Applications - Computer Vision

• 2 DBNs.• 1st DBN *encodes* data into vector of 10-30

numbers = Pre-training.• 2nd DBN decodes data into original state.• Backprop only happens on 2nd DBN• 2nd is the fine-tuning stage (reconstruction

entropy).• Reduces documents or images to compact

vectors .• Useful in search, QA and information

retrieval.

Deep Autoencoder

Page 16: Deep Learning and its Applications - Computer Vision

Deep Autoencoder Architecture

Page 17: Deep Learning and its Applications - Computer Vision

Image Search Results

Page 18: Deep Learning and its Applications - Computer Vision

• Top-down & hierarchical rather than feed-forward (DBNs).

• Handles sequence-based classification, windows of several events, entire scenes (multiple objects).

• Features themselves are vectors. • A tensor = a multi-dimensional matrix, or multiple

matrices of the same size.

Recursive Neural Tensor Net

Page 19: Deep Learning and its Applications - Computer Vision

RNTNs & Scene Composition