The presentation has many use cases covering the following Image classification: "The process of identifying and detecting an object or a feature in a digital image or video," the report states. In retail, deep learning models "quickly scan and analyze in-store imagery to intuitively determine inventory movement."
Voice recognition: "The ability to receive and interpret dictation or to understand and carry out spoken commands. Models are able to convert captured voice commands to text and then use natural language processing to understand what is being said and in what context." In transportation, deep learning "uses voice commands to enable drivers to make phone calls and adjust internal controls - all without taking their hands off the steering wheel."
Anomaly detection: "Deep learning technique strives to recognize abnormal patterns which don't match the behaviors expected for a particular system, out of millions of different transactions. These applications can lead to the discovery of an attack on financial networks, fraud detection in insurance filings or credit card purchases, even isolating sensor data in industrial facilities signifying a safety issue."
Recommendation engines: "Analyze user actions in order to provide recommendations based on user behavior."
Sentiment analysis: "Leverages deep learning-heavy techniques such as natural language processing, text analysis, and computational linguistics to gain clear insight into customer opinion, understanding of consumer sentiment, and measuring the impact of marketing strategies."
Video analysis: "Process and evaluate vast streams of video footage for a range of tasks including threat detection, which can be used in airport security, banks, and sporting events."
Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
Deep Learning Image Processing Applications in the Enterprise
1. Deep Learning Image
Processing Applications
in the Enterprise
Clarisse Taaffe-Hedglin
Executive IT Architect
IBM Garage
IBM Systems
clarisse@us.ibm.com
2. Agenda
The AI Ladder and Lifecycle
Deep Learning Use cases
AI at Scale Themes
Infrastructure Considerations
3. “AI is the
fastest-growing
workload”*
3*Forrester Research Inc. “AI Deep Learning Workloads Demand a New Approach to Infrastructure”, by
Mike Gualtieri, Christopher Voce, Srividya Sridharan, Michele Goetz, Renee Taylor, May 4, 2018.
5. Machine Learning Definition
5
“…an application of artificial
intelligence (AI) that provides systems
the ability to automatically learn and
improve from experience without
being explicitly programmed.”
6. 6
Enterprise Analytics Modernization: From Data to Actions
010101010101010111100010011001010111
0000000000010101010100000000000 111101011
11000 000000000000 111111 010101 101010 10101010100
Prescriptive
What should
we do ?
Descriptive
What Has
Happened?
Cognitive
Learn
Dynamically
Predictive
What Will
Happen?
ACTIONDATA
HUMAN INPUTS
<
< >
< >
>
>
7. Three broad categories of AI Use Cases
“Structured” Data Use Cases
Computer Vision Use Cases
- Big Data (Rows and Columns)
- GPU Servers
- Available AI Software
More Accuracy !
This is sort of “Magic”
- a deep learning Model is trained to detect and classify objects
Natural Language Processing Use Cases
- A Model learns to read and hear and “understand” language
8. Predict a
Future Event
Segment Data
/ Detect
Anomalies
Determine
optimal
quantity,
price,
resource
allocation, or
best action
Understand
Past Activity
Discover
Insights in
Content
(text, images,
video)
Interact in
Natural
Language
Forecast
and Budget
based on
past activity
Supervised Unsupervised
Predictive: What will happen? Prescriptive:
What should
we do?
Descriptive:
What
happened?
Planning:
What is our
Plan?
NLPDeep Learning
Supervised
Common Patterns of Analytics
Solving challenges with Data and AI
will utilize a combination of these analytics patterns
9. AI and Autonomous Machine
Learning will help revolutionized
every single industry making us
more productive and efficient
to do things that today are
impossible to do.
Art of the Possible
10. Organizations are adopting
AI to solve business problems
Fraud Safety, inspection and
process improvement
Defense and security
11. MEDIA/ENTERTAINMENT
RETAIL
Recomendation
engines, Precision
marketing
AGRICULTURE
Crop yield, Plant
disease, remote
sensing
LIFE SCIENCES
Sequence
Analysis,
Radiology
UTILITIES
Smart Meter analysis,
Capacity planning
$
FINANCIAL SERVICES
Risk analysis
Fraud detection
CUSTOMER SERVICE
Chatbots, Helpdesk
Automated
Expenses
LAW & DEFENSE
Threat analysis -
social media
monitoring
RESEARCH
Physics Modeling
Simulation optimization
HEALTH CARE
Patient sensors,
monitoring, EHRs
TRANSPORTATION
Optimal traffic
flows, Route
planning
CONSUMER GOODS
Sentiment
analysis
Advertising
effectiveness
OIL & GAS
Exploration,
sensor analysis
AUTOMOTIVE
ADAS,
Maintenance
MANUFACTURING
Line
inspection,
Defect analysis
Addressable Markets and Fields for AI
12. “AI is the
fastest-growing
workload”*
12*Forrester Research Inc. “AI Deep Learning Workloads Demand a New Approach to Infrastructure”, by
Mike Gualtieri, Christopher Voce, Srividya Sridharan, Michele Goetz, Renee Taylor, May 4, 2018.
13. COLLECT - Make data simple and accessible
ORGANIZE - Create a trusted analytics foundation
ANALYZE - Scale AI everywhere with trust & transparency
Data of every type, regardless of
where it lives
MODERNIZE
your data estate for an
AI and multicloud world
INFUSE – Operationalize AI across business processes
The AI Ladder
A prescriptive approach to accelerating the journey to AI
13
AI
AI-optimized systems
infrastructure
14. Unstructured, Landing, Exploration and Archive
Operational Data
Real-time Data Processing & Analytics
Transaction and
application data
Machine,
sensor data
Enterprise
content
Image, geospatial,
video
Social data
Third-party data
Information Integration & Governance
Data is Prerequisite to AI
Risk, Fraud
Chat bots,
personal
assistants
Supply Chain
Optimization
Dynamic
Pricing,
Recommenders
Behavior
Modeling
Vision,
Autonomous
Systems
17. Metadata-Fueled Data Analysis
Large Scale Data Ingest
• Scan records at high speed
• Live event notifications
• Capture system-level tags
• Automatic indexing
Business-Oriented
Data Mapping
• Custom data tagging
• Content-inspection via APIs
• Policy-driven workflows
Data Activation
• Data movement via APIs
• Extensible architecture
• Solution Blueprints
Data Visualization
• Query billions of records
in seconds
• Multi-faceted search
• Drilldown dashboard
• Customizable reports
18. AI Model Development Workflow
•Data preparation, cleaning, labelling
•Model development environment
•Runtime environment
•Train, deploy and manage models
•Business KPI and production metrics
•Explainability and fairness
Data Engineering and Data Science Team IT Operations Team
19. Data Science Exploration
to Production
Use Case Exploration
Data Science Model Build
Use Case Deployment in Production
Requires solution architecture
Deploy
Source: https://papers.nips.cc/paper/5656-hidden-technical-debt-in-machine-learning-systems.pdf
Use Case Exploration
Data Science Model Build
Security, Privacy and Governance
20. • #1 Model
Quality
• Not enough
knowledge about the
problem to build a
good model
• #2 System
Usage
• Typical optimizations
are limited to serial
data collection
• #3 Complexity
• Typical optimizations
do not work in high
dimensions
• #4 Trust
• Typical optimizations
do not explain their
logic to the user
Designing Models Driven By User Desires
21.
22. AI Use Case: Automate diagnostics
to increase productivity
DIAGNOSTICS
Faster results with higher accuracy
can be achieved with an image
processing system designed to
• address workflow burdens,
• data governance challenges, and
• analysis challenges
with the goal of reducing
• false negative rates in imaging
diagnostics and in clinical settings,
• patient risk and medical legal risk.
23. Examples of Medical Imaging Applications
s.
3D-UNet segmentation models with
higher resolution images allows for
learning and labeling finer details
and structures of brain tumors.
https://developer.ibm.com/linuxonpower/2018/07/27/tensorflow-
large-model-support-case-study-3d-image-segmentation/
Automatic skin lesion image
analysis for melanoma detection
with Memorial Sloan Kettering
(MSK-CC)
Diagnosis of blood-based with
characterization of patient blood
samples to detect and classify
blood cell subtypes
DIAGNOSTICS
24. Optimizing Medical Imaging
Enhance image identification with deep learning
to assist physicians and benefit patients
1300 MRI images trained by IBM
Power Systems and IBM Storage in
just two hours, compared to forty
hours on traditional architectures
20x faster
DIAGNOSTICS
27. Accelerated workflow
uses fewer calculations
to achieve orders of
magnitude resolution
increase
AI Use Case: Molecular Modeling
Achieves human level
performance in days
instead of months.
Force Field Tuning
Intelligent Phase Diagram Exploration
MOLECULAR SIMULATION
28. 28
• Advances in instrument design, sample
preprocessing and mathematical
methods have enabled high volume
throughput imaging at atomic scale.
• Cryogenic electron microscopes
generate an average of 5 TB of image
data per day
BIOMOLECULAR STRUCTURE
AI Use Case: Massive data sets
require massive processing capability
29. Accelerating Cryo-EM Imaging Analysis
Reduced time-to-completion for high resolution image
analysis jobs while increasing resource utilization
Using IBM AC922 cluster, more than 100 cryo-EM
high resolution image workload analysis jobs running
in parallel on Satori cluster
100+
BIOMOLECULAR STRUCTURE
30. Traditional infrastructure isn’t
suited for AI workloads
Systems don't easily scale
to meet demand
Processor not optimized for
AI workloads
The wrong infrastructure puts AI at risk.
Data pipeline too slow, causing
bottleneck effect
31. Common AI Data Considerations
Data Compute
Legacy Data
Stores
IoT, Mobile
& Sensors
Collaboration
Partners
New Data
Ingest InferenceTrainingPreparation
Iterative Model training to improve accuracy
Champion
Challenge
r
-”Data Center”
- At Edge
Trained
Model
§ Ease to Massively Scale
§ High Performance
§ Tiered / Archive
§ Secure
§ High Performance
§ Metadata Tagging
§ Single Name Space
Low Latency
Dev & Inference Stack
- Open Source
- Stable and Supported
- Auditable
Productivity
Performance
Robustness
Considerations
32. Infrastructure
Demands for AI
Equipped for volumes of data
Flexible storage for a range
of data demands
Versatile, power-efficient data
center accelerators
Advanced I/O for minimal latency
Scalability and distributed
data center capability
Inference
Powerful data center
accelerators with coherence
Advanced I/O for high
bandwidth and low latency
Proven scalability
Training
Equipped for volumes of data
34. Inferencing Considerations
Real-Time (vs Batch): Many AI applications
have response times in milli-seconds and in
many cases have 100K+ IOT events per
second (Latency, Latency, Latency)
Scalability: Ability to scale inference engine
and manage infrastructure
Data Pipeline: The data that is feed into
models has to be cleaned and structured to
produce accurate results
Security: Applications running AI models in
the field and back-offices
Multi-Tenancy: Multiple business
applications leveraging shared
infrastructure, Multiple Models per Business
Application
Tools Proliferation: Analytics, Data/Object
Tagging, Model Training and Inferencing
Model Management: Continuous
Training/Re-Training of Models, AI-DevOps,
Ease of Deployment
Transparency: Ability to explain decisions
A
C
C
U
R
A
C
Y
Transaction integration
Huge Scale
As-a-Service offering
Inference Data Center or In-Cloud
Multi-Tenancy
Low latency
Data movement considerations
Near Edge Inferencing
On-prem or In-Cloud
Inference at Edge
On-prem/device
Stand alone device
Low latency
Data movement considerations
Typical AI Inferencing Scenarios
37. OpenPOWER is a technical community
dedicated to expanding the the IBM Power architecture ecosystem
https://github.com/open-ce
Open-CE
Minimize time to value for
foundational ML/DL packages
Provide a flexible source-to-image
solution to provide a complete and
customizable AI environment.
38.
39. Data Data Data
Microservices Containerized Workloads Multicloud Provisioning
Public Cloud
On-prem
ises
An architecture of loosely coupled
data services, easily refactored to
create containerized workloads
Stand-alone workloads composed of
microservices & data that are flexibly
deployed, orchestrated and managed
Agile provisioning of containerized
workloads in multicloud environments
and consumption of cloud services
Cloud Native Platforms
Agility Efficiency Cost Savings
IBM Cloud Pak for Data
45. Provision Faster Scale Affordably Maximize Uptime
• Provision SAP HANA instances
faster with built-in virtualization
• Easily make capacity changes
• Simplify management consolidating
HANA instances
• Minimize infrastructure with scale
up environment
• Granular capacity allocation
• Share and optimize CPU allocation
• Capacity on Demand
• Ranked most reliable server for
over a decade1
• Zero impact planned
maintenance with LPM
• Virtual persistent memory for
faster restart and shutdown
1. ITIC 2018 Global Server Hardware, Server OS Reliability Survey Mid-Year Update. The highest uptime of 99.9996% is
calculated based on 2.0 minutes/server/annum unplanned downtime of any non-mainframe Linux platforms
Your smart choice to run SAP HANA
IBM Power Systems
45
46. What is SAS Viya?
A cloud-enabled, in-memory analytics engine
– Provides quick, accurate and reliable analytical insights.
– Elastic, scalable and fault-tolerant processing addresses the complex analytical challenges of
today
– Effortlessly scaling for the future.
SAS Viya provides:
– Faster processing for huge amounts of data and the most complex analytics,
– Including machine learning, deep learning and artificial intelligence
– Standardized code base that supports programming in SAS and other languages, like Python,
R, Java and Lua.
– Support for cloud, on-site or hybrid environments.
– It deploys seamlessly to any infrastructure or application ecosystem.
Page
46
47. 47Page
SAS 9.4 & SAS Viya
Similarities/ Differences/ Relationships
SAS® 9.4
– Discover insights, manage data and
make analytics approachable. Legacy
SAS.
SAS Viya
– Cloud-enabled, in-memory analytics
engine that provides quick, accurate
and reliable analytical insights.
They compliment each other –
not a direct replacement
SAS Visual
Analytics
SAS
Report
Viewer
48. 48
SAS Visual Analytics
SAS high-performance technologies accelerate analytic computations derive value from massive
amounts of data
49. Eliminate bottlenecks
• Generate insights on time, every
time by scaling on demand
• Easily allocate precise capacity at
the push of a button
• Simplify management with co-
located workloads in same system
• Optimize resource utilization
Drive agility Reduce risk
• Reduce risk with #1 ranked
systems in reliability
• Zero impact planned
downtime with Live Partition
Mobility
• Eliminate bottlenecks with the
industry leading throughput
• 2x I/O and 1.8x memory
bandwidth vs compared x86
platforms
1. ITIC 2018 Global Server Hardware, Server OS Reliability Survey Mid-Year Update. The highest uptime of 99.9996% is
calculated based on 2.0 minutes/server/annum unplanned downtime of any non-mainframe Linux platforms
Accelerate insights from SAS solutions
with
IBM Power Systems
49Page
50. Best Practice Approach:
Think Solutions !
Gaining insights with Machine Learning/Deep
Learning requires a flexible end to end
solution first approach
Focus on solving problems and use cases
Data is a pre-requisite
ML/DL is just a piece of an overall workflow
Infrastructure matters
Establish trusted collaborations, partners
In Summary