SlideShare a Scribd company logo
1 of 30
Download to read offline
TouchNet preview
by Jason Toy
TouchNet
• A touch simulator and dataset
of touchable objects to teach
AIs how to interact with their
environments via motor-
sensory neurons
history
• idea around 2015
• started coding September 2017
• initial release 2017 Q4
why TouchNet?
why TouchNet?
• On Intelligence
• future predicting machines
• neuroscience based principles
why TouchNet?
• Helen Keller
• intelligence through fingers
• consciousness “started” from I
Helen Keller
• “Every object is associated in my mind with tactual
qualities which, combined in countless ways, give me
a sense of power, of beauty.”
• “..with the fingers I see only a very small portion of a
surface, and that I must pass my hand continually over
it before my touch grasps the whole. It is still more
important, however, to remember that my imagination
is not tethered to certain points, locations, and
distances. It puts all the parts together simultaneously
as if it saw or knew instead of feeling them.”
where is AI now?
• images, CNNs, LSTMs, GANs
• most business focus on static models; some
work in RL
• where is time?
• motor sensory neurons?
• where is the agent compared to the world?
- John Locke, philosophy of mind
“An unknown, but potentially large, fraction of
animal and human intelligence is a direct
consequence of the perceptual and physical
richness of our environment, and is unlikely to
arise without it.”
TouchNet consists of 4 parts
• dataset of 3D objects to touch and interact with
• an evolving simulation
• a simple extendable API
• benchmarks to measure and track progress
Dataset
Dataset
• procedurally generated 3D objects
• common objects: cups, plates, bowls, etc
• starting with tool objects
• cifar100 …. imagenet
Simulator
Simulators and physics
• need a real physics engine
• needs to be fully programmable & modifiable
• tested all of them: MuJoCo, DeepMind Lab,
Roboschool, project Malmo, VizDoom,Gazebo,
bullet , Processing
Simulators and physics
• bullet and pybullet are a secret weapon!
• robust physics engine used for many applications
• super easy to install
• movies: Shrek, Hancock,Sherlock Holmes….
• games: Toy Story 3, Grand Theft Auto 4,Hot
Wheels…
pybullet
• rigid and soft body simulations
• collision detection
• works with unity, blender,CUDA,OpenCL,…..
• developed by Erwin Coumans & community since
~2003
• now at Google and used for RL & Robotics
• http://pybullet.org
What we want to simulate
What we have
simulation demo
coevolving simulator & agent
• index finger with merkel touch sensors
• all sensor types in finger
• all fingers
• full arm
• full body?
API
• simple to get started and use
• easy to install
• modifiable and customizable
reinforcement learning api
• based off OpenAI’s gym and universe API
• can plug in almost any code for reinforcement learning
OpenAI’s Universe API
Same API
Benchmarks
“Blind” objection
classification
• goal: classify the object
• CNN and actor critic DNN models trained
concurrently
• pytorch and tensorflow versions
• 4% accuracy, please beat mine!
“blind” classification
code walk through
Benchmarks
• HTM version?
• more benchmarks in the future e.g:
• moving objects, differentiating textures,…..
Help!
• very ambitious project, too big for me
• 100% open source
• I do a little of everything ( suck at everything? )
• looking for contributors and experts in
physics,math, neuroscience, ml
Jason Toy
jtoy@jtoy.net
twitter.com/jtoy
Thank you!
Questions and Feedback
http://github.com/jtoy/touchnet

More Related Content

What's hot

What's hot (11)

Evo mouse
Evo mouseEvo mouse
Evo mouse
 
Sixth Sense technology
Sixth Sense technologySixth Sense technology
Sixth Sense technology
 
Sixthsense technology
Sixthsense technologySixthsense technology
Sixthsense technology
 
Building VR Applications For Google Cardboard
Building VR Applications For Google CardboardBuilding VR Applications For Google Cardboard
Building VR Applications For Google Cardboard
 
Getting started with Unity and AR/VR for the .NET developer - October 2020
Getting started with Unity and AR/VR for the .NET developer - October 2020Getting started with Unity and AR/VR for the .NET developer - October 2020
Getting started with Unity and AR/VR for the .NET developer - October 2020
 
Presentation on Google Tango By Atharva Jawalkar
Presentation on Google Tango By Atharva Jawalkar Presentation on Google Tango By Atharva Jawalkar
Presentation on Google Tango By Atharva Jawalkar
 
Evo mouse
Evo mouseEvo mouse
Evo mouse
 
2016 AR Summer School - Lecture 5
2016 AR Summer School - Lecture 52016 AR Summer School - Lecture 5
2016 AR Summer School - Lecture 5
 
Getting started with unity and AR/VR for the .net developer - DDD North 2020
Getting started with unity and AR/VR for the .net developer - DDD North 2020Getting started with unity and AR/VR for the .net developer - DDD North 2020
Getting started with unity and AR/VR for the .net developer - DDD North 2020
 
Virtual Reality
Virtual RealityVirtual Reality
Virtual Reality
 
Interaction Design
Interaction DesignInteraction Design
Interaction Design
 

Similar to TouchNet preview at Numenta

Chris Wren (WrenAR LLC): Beyond the Hands
Chris Wren (WrenAR LLC): Beyond the HandsChris Wren (WrenAR LLC): Beyond the Hands
Chris Wren (WrenAR LLC): Beyond the Hands
AugmentedWorldExpo
 

Similar to TouchNet preview at Numenta (20)

Closing the gap between neuroscience and machine learning
Closing the gap between neuroscience and machine learningClosing the gap between neuroscience and machine learning
Closing the gap between neuroscience and machine learning
 
Microsoft Hololens
Microsoft HololensMicrosoft Hololens
Microsoft Hololens
 
Tactile brush
Tactile brushTactile brush
Tactile brush
 
Empathic Computing
Empathic ComputingEmpathic Computing
Empathic Computing
 
5G and the Invisible Interface
5G and the Invisible Interface5G and the Invisible Interface
5G and the Invisible Interface
 
HoloLens
HoloLensHoloLens
HoloLens
 
QS Tools for Emotions and Communication
QS Tools for Emotions and CommunicationQS Tools for Emotions and Communication
QS Tools for Emotions and Communication
 
Designing agents as if people mattered
Designing agents as if people matteredDesigning agents as if people mattered
Designing agents as if people mattered
 
Sixth sense technology ppt
Sixth sense technology pptSixth sense technology ppt
Sixth sense technology ppt
 
Microsoft hololens
Microsoft hololensMicrosoft hololens
Microsoft hololens
 
Marek Rosa - Inventing General Artificial Intelligence: A Vision and Methodology
Marek Rosa - Inventing General Artificial Intelligence: A Vision and MethodologyMarek Rosa - Inventing General Artificial Intelligence: A Vision and Methodology
Marek Rosa - Inventing General Artificial Intelligence: A Vision and Methodology
 
Microsoft hololens
Microsoft hololensMicrosoft hololens
Microsoft hololens
 
Artificial Intelligence by Sushil Louis
Artificial Intelligence by Sushil LouisArtificial Intelligence by Sushil Louis
Artificial Intelligence by Sushil Louis
 
Introduction of machine learning.pptx
Introduction of machine learning.pptxIntroduction of machine learning.pptx
Introduction of machine learning.pptx
 
Google io 2015, Trip Report
Google io 2015, Trip ReportGoogle io 2015, Trip Report
Google io 2015, Trip Report
 
Machine learning
Machine learningMachine learning
Machine learning
 
Chris Wren (WrenAR LLC): Beyond the Hands
Chris Wren (WrenAR LLC): Beyond the HandsChris Wren (WrenAR LLC): Beyond the Hands
Chris Wren (WrenAR LLC): Beyond the Hands
 
Ai lecture1 final
Ai lecture1 finalAi lecture1 final
Ai lecture1 final
 
Deep learning introduction
Deep learning introductionDeep learning introduction
Deep learning introduction
 
What's Next: Cloudy with a chance of AI – Deep Learning
What's Next: Cloudy with a chance of AI – Deep LearningWhat's Next: Cloudy with a chance of AI – Deep Learning
What's Next: Cloudy with a chance of AI – Deep Learning
 

More from jtoy (7)

Rga somatic hackday
Rga somatic hackdayRga somatic hackday
Rga somatic hackday
 
Experiential Marketing and Deep Learning
Experiential Marketing and Deep LearningExperiential Marketing and Deep Learning
Experiential Marketing and Deep Learning
 
Intro to myself and SFPC
Intro to myself and SFPCIntro to myself and SFPC
Intro to myself and SFPC
 
Tensorflow ruby
Tensorflow rubyTensorflow ruby
Tensorflow ruby
 
machine learning and art hack day
machine learning and art hack daymachine learning and art hack day
machine learning and art hack day
 
a tour of several popular tensorflow models
a tour of several popular tensorflow modelsa tour of several popular tensorflow models
a tour of several popular tensorflow models
 
Bm packing list
Bm packing listBm packing list
Bm packing list
 

Recently uploaded

Conjugation, transduction and transformation
Conjugation, transduction and transformationConjugation, transduction and transformation
Conjugation, transduction and transformation
Areesha Ahmad
 
Seismic Method Estimate velocity from seismic data.pptx
Seismic Method Estimate velocity from seismic  data.pptxSeismic Method Estimate velocity from seismic  data.pptx
Seismic Method Estimate velocity from seismic data.pptx
AlMamun560346
 
Pests of mustard_Identification_Management_Dr.UPR.pdf
Pests of mustard_Identification_Management_Dr.UPR.pdfPests of mustard_Identification_Management_Dr.UPR.pdf
Pests of mustard_Identification_Management_Dr.UPR.pdf
PirithiRaju
 

Recently uploaded (20)

Conjugation, transduction and transformation
Conjugation, transduction and transformationConjugation, transduction and transformation
Conjugation, transduction and transformation
 
GBSN - Microbiology (Unit 1)
GBSN - Microbiology (Unit 1)GBSN - Microbiology (Unit 1)
GBSN - Microbiology (Unit 1)
 
COST ESTIMATION FOR A RESEARCH PROJECT.pptx
COST ESTIMATION FOR A RESEARCH PROJECT.pptxCOST ESTIMATION FOR A RESEARCH PROJECT.pptx
COST ESTIMATION FOR A RESEARCH PROJECT.pptx
 
High Class Escorts in Hyderabad ₹7.5k Pick Up & Drop With Cash Payment 969456...
High Class Escorts in Hyderabad ₹7.5k Pick Up & Drop With Cash Payment 969456...High Class Escorts in Hyderabad ₹7.5k Pick Up & Drop With Cash Payment 969456...
High Class Escorts in Hyderabad ₹7.5k Pick Up & Drop With Cash Payment 969456...
 
Clean In Place(CIP).pptx .
Clean In Place(CIP).pptx                 .Clean In Place(CIP).pptx                 .
Clean In Place(CIP).pptx .
 
High Profile 🔝 8250077686 📞 Call Girls Service in GTB Nagar🍑
High Profile 🔝 8250077686 📞 Call Girls Service in GTB Nagar🍑High Profile 🔝 8250077686 📞 Call Girls Service in GTB Nagar🍑
High Profile 🔝 8250077686 📞 Call Girls Service in GTB Nagar🍑
 
COMPUTING ANTI-DERIVATIVES (Integration by SUBSTITUTION)
COMPUTING ANTI-DERIVATIVES(Integration by SUBSTITUTION)COMPUTING ANTI-DERIVATIVES(Integration by SUBSTITUTION)
COMPUTING ANTI-DERIVATIVES (Integration by SUBSTITUTION)
 
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
 
Factory Acceptance Test( FAT).pptx .
Factory Acceptance Test( FAT).pptx       .Factory Acceptance Test( FAT).pptx       .
Factory Acceptance Test( FAT).pptx .
 
Proteomics: types, protein profiling steps etc.
Proteomics: types, protein profiling steps etc.Proteomics: types, protein profiling steps etc.
Proteomics: types, protein profiling steps etc.
 
Zoology 5th semester notes( Sumit_yadav).pdf
Zoology 5th semester notes( Sumit_yadav).pdfZoology 5th semester notes( Sumit_yadav).pdf
Zoology 5th semester notes( Sumit_yadav).pdf
 
Site Acceptance Test .
Site Acceptance Test                    .Site Acceptance Test                    .
Site Acceptance Test .
 
PSYCHOSOCIAL NEEDS. in nursing II sem pptx
PSYCHOSOCIAL NEEDS. in nursing II sem pptxPSYCHOSOCIAL NEEDS. in nursing II sem pptx
PSYCHOSOCIAL NEEDS. in nursing II sem pptx
 
Call Girls Alandi Call Me 7737669865 Budget Friendly No Advance Booking
Call Girls Alandi Call Me 7737669865 Budget Friendly No Advance BookingCall Girls Alandi Call Me 7737669865 Budget Friendly No Advance Booking
Call Girls Alandi Call Me 7737669865 Budget Friendly No Advance Booking
 
Unit5-Cloud.pptx for lpu course cse121 o
Unit5-Cloud.pptx for lpu course cse121 oUnit5-Cloud.pptx for lpu course cse121 o
Unit5-Cloud.pptx for lpu course cse121 o
 
Connaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verified
Connaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verifiedConnaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verified
Connaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verified
 
GBSN - Microbiology (Unit 3)
GBSN - Microbiology (Unit 3)GBSN - Microbiology (Unit 3)
GBSN - Microbiology (Unit 3)
 
Seismic Method Estimate velocity from seismic data.pptx
Seismic Method Estimate velocity from seismic  data.pptxSeismic Method Estimate velocity from seismic  data.pptx
Seismic Method Estimate velocity from seismic data.pptx
 
9999266834 Call Girls In Noida Sector 22 (Delhi) Call Girl Service
9999266834 Call Girls In Noida Sector 22 (Delhi) Call Girl Service9999266834 Call Girls In Noida Sector 22 (Delhi) Call Girl Service
9999266834 Call Girls In Noida Sector 22 (Delhi) Call Girl Service
 
Pests of mustard_Identification_Management_Dr.UPR.pdf
Pests of mustard_Identification_Management_Dr.UPR.pdfPests of mustard_Identification_Management_Dr.UPR.pdf
Pests of mustard_Identification_Management_Dr.UPR.pdf
 

TouchNet preview at Numenta

  • 2. TouchNet • A touch simulator and dataset of touchable objects to teach AIs how to interact with their environments via motor- sensory neurons
  • 3. history • idea around 2015 • started coding September 2017 • initial release 2017 Q4
  • 5. why TouchNet? • On Intelligence • future predicting machines • neuroscience based principles
  • 6. why TouchNet? • Helen Keller • intelligence through fingers • consciousness “started” from I
  • 7. Helen Keller • “Every object is associated in my mind with tactual qualities which, combined in countless ways, give me a sense of power, of beauty.” • “..with the fingers I see only a very small portion of a surface, and that I must pass my hand continually over it before my touch grasps the whole. It is still more important, however, to remember that my imagination is not tethered to certain points, locations, and distances. It puts all the parts together simultaneously as if it saw or knew instead of feeling them.”
  • 8. where is AI now? • images, CNNs, LSTMs, GANs • most business focus on static models; some work in RL • where is time? • motor sensory neurons? • where is the agent compared to the world?
  • 9. - John Locke, philosophy of mind “An unknown, but potentially large, fraction of animal and human intelligence is a direct consequence of the perceptual and physical richness of our environment, and is unlikely to arise without it.”
  • 10. TouchNet consists of 4 parts • dataset of 3D objects to touch and interact with • an evolving simulation • a simple extendable API • benchmarks to measure and track progress
  • 12. Dataset • procedurally generated 3D objects • common objects: cups, plates, bowls, etc • starting with tool objects • cifar100 …. imagenet
  • 14. Simulators and physics • need a real physics engine • needs to be fully programmable & modifiable • tested all of them: MuJoCo, DeepMind Lab, Roboschool, project Malmo, VizDoom,Gazebo, bullet , Processing
  • 15. Simulators and physics • bullet and pybullet are a secret weapon! • robust physics engine used for many applications • super easy to install • movies: Shrek, Hancock,Sherlock Holmes…. • games: Toy Story 3, Grand Theft Auto 4,Hot Wheels…
  • 16. pybullet • rigid and soft body simulations • collision detection • works with unity, blender,CUDA,OpenCL,….. • developed by Erwin Coumans & community since ~2003 • now at Google and used for RL & Robotics • http://pybullet.org
  • 17. What we want to simulate
  • 20. coevolving simulator & agent • index finger with merkel touch sensors • all sensor types in finger • all fingers • full arm • full body?
  • 21. API • simple to get started and use • easy to install • modifiable and customizable
  • 22. reinforcement learning api • based off OpenAI’s gym and universe API • can plug in almost any code for reinforcement learning
  • 26. “Blind” objection classification • goal: classify the object • CNN and actor critic DNN models trained concurrently • pytorch and tensorflow versions • 4% accuracy, please beat mine!
  • 28. Benchmarks • HTM version? • more benchmarks in the future e.g: • moving objects, differentiating textures,…..
  • 29. Help! • very ambitious project, too big for me • 100% open source • I do a little of everything ( suck at everything? ) • looking for contributors and experts in physics,math, neuroscience, ml
  • 30. Jason Toy jtoy@jtoy.net twitter.com/jtoy Thank you! Questions and Feedback http://github.com/jtoy/touchnet