Deep Learning At the edge of the network Hugo Latapie / Enzo - - PowerPoint PPT Presentation

▶
deep learning
SMART_READER_LITE
LIVE PREVIEW

Deep Learning At the edge of the network Hugo Latapie / Enzo - - PowerPoint PPT Presentation

Apr 16, 2023 166 likes •361 views

Deep Learning At the edge of the network Hugo Latapie / Enzo Fenoglio Santosh Pandey, JP Vasseur, Rob Liston, Dan Tan, Xiaoqing Zhu, John Apostolopoulos with the newly formed IoE Video Analytics virtual team HPC Since 2011 Audio/video

slide-1
SLIDE 1

Hugo Latapie / Enzo Fenoglio Santosh Pandey, JP Vasseur, Rob Liston, Dan Tan, Xiaoqing Zhu, John Apostolopoulos with the newly formed IoE Video Analytics virtual team

At the edge of the network

Deep Learning

slide-2
SLIDE 2
  • Audio/video metadata extraction
  • Multi-modal TV UX
  • Encrypted Network Traffic

Classification

  • Video compression

HPC Since 2011

slide-3
SLIDE 3

Public/Private Internet

DLA DLA

Branch 1

SLN Architecture

Internet Branch 2

Controller

ISE

SCA Threat Intel Threat Grid, OpennDNS, WBRS, ... Other TI feeds

  • Sensing (knowledge): granular data

collection with knowledge extraction from NetFlow but also Deep Packet Inspection

  • n control and data plane & local states
  • Machine Learning: real-time embedded

behavioral modeling and anomaly detection

  • Control: autonomous embedded control,

advanced networking control (police, shaper, recoloring, redirect, ...) DLA

  • Orchestration of Distributed Learning

Agents (DLAs)

  • Advanced Visualization of anomalies
  • Centralized policy for mitigation
  • Interaction with other security components

such as ISE and Threat Intelligence Feeds

  • North bound API to SIEM/Database (e.g.

Splunk) using CEF/CIM format

  • Evaluation of anomaly relevancy

SCA

slide-4
SLIDE 4
slide-5
SLIDE 5
slide-6
SLIDE 6
slide-7
SLIDE 7

Evolution of location use-cases

  • Healthcare

Asset management & wayfinding

  • Retail

Engage shopper in aisle & deliver proximity-based offers

  • Museum

Enabling the Digital docent

  • Office

Wayfinding & workspace optimization

93%

slide-8
SLIDE 8
  • State of the art:
  • 5-7m accuracy
  • Multilateration on WiFi Client based on RSSI at multiple APs
  • Cisco-Hyperlocation:
  • 1m accuracy
  • Increase accuracy & reliability
  • AP connected clients
  • Not applicable to non-connected/probing clients or tags
  • Add Angle-of-Arrival in addition to RSSI
  • Introduce enhanced location with 1AP

Technology – network based location

  • RSSI

Receive Signal Strength Indicator

  • ToF

Time of Flight

  • DRTT Differential Round-trip-time
slide-9
SLIDE 9

How Hyperlocation Works

Built on Cisco Unified Access

CMX Analytics

Controller (Virtual/Physical) MSE (Virtual/Physical) Analytics UI

LOCATION DATA CMX Connect APPLICATION DATA

Mobile Application Server

Depending on Application Layer

HyperLocation Access Points

slide-10
SLIDE 10

MVP 1 -- Counting

slide-11
SLIDE 11

Scalable video analytics in crowded scenes

Phase 2

  • General DL pipeline based on nvidia-docker
  • verlay networking and swarm with support

for thousands of nodes and 10’s of thousands of containers

  • High performance video pipeline based on

gstreamer

Phase 1

slide-12
SLIDE 12
  • Tracking involves extracting the spatio-temporal history of each

person in a scene solving the assignment problem using Jonker- Volgenant algorithm

  • We use a Convolutional Network (CNN) + LSTM
  • Tracking estimation is performed by multiple Kalman filters assigned

to each tracklet

Tracking Overview CNN/ LSTM KF/JV

Input RGB frames Localization Estimation Tracking estimation

slide-13
SLIDE 13

Back to data fusion

slide-14
SLIDE 14

An intersection of technology and trends:

  • Recent camera sensor technologies offer advanced capabilities
  • Advanced cameras + analytics create powerful IoT sensors
  • Visual analytics and data fusion fit naturally into the fog architecture

The industry is moving towards scalable, flexible, distributed analytics platforms

Machine Learning from End Point to Cloud

Heterogeneous Compute Platform CPU + GPU/FPGA

IoT Cloud Software

Advanced Cameras

slide-15
SLIDE 15
  • 1. Better Understanding of the public space
  • 2. Optimization of energy management in

public buildings CDP, MERAKI wifi, CMX, LoRaWAN, cameras, video analytics, Energy / Asset Management

2 innovative pilots proposed by the City: outdoor/indoor

slide-16
SLIDE 16
slide-17
SLIDE 17

Cisco Data Center Portfolio and NVIDIA

  • Virtual workstation for high-end graphics

applications

  • Cisco UCS C240 M4 Rack servers
  • Support NVIDIA GRID 1.0 and 2.0 with K1/K2 and M60 cards
  • Support for Magma ExpressBox for higher density
  • Cisco UCS B200 M4 Blade servers
  • Recently introduced NVIDIA M6 MXM support on Blade server
  • Cisco HyperFlex –2nd generation HyperConverged

platform

  • Phase 1: K1/K2 support
  • Phase 2: M6/M60 support
  • Deep Learning and HPC
  • Cisco UCS C240 Rack servers with TESLA K80
slide-18
SLIDE 18