SlideShare a Scribd company logo
Continual Learning: Another Step
TowardsTruly Intelligent Machines
Introduction Meetup @ Numenta
16-09-2019
Vincenzo Lomonaco
vincenzo.lomonaco@unibo.it
Postdoctoral Researcher @ University of Bologna
Supervisor: Davide Maltoni
About me
• Post-Doc @ University of Bologna
• Research Affiliate @ AI Labs
• Teaching Assistant of the courses
Machine Learning and Computer
Architectures @ UniBo
• Author andTechnical reviewer of the
online course Deep Learning with R and
book R Deep Learning Essentials.
• Co-Founder and President of
ContinualAI.org
• Co-Founder and Board Member of Data
Science Bologna and AIforPeople.org
What’s ContinualAI?
• ContinualAI is a non-profit research organization and
the largest research community on Continual Learning
for AI.
• It counts more than 550+ members in 17 different
time-zones and from top-notch research institutions.
• Learn more about ContinualAI at www.continualai.org
ContinualAI Board Members and Advisors
Machine Intelligence @ BioLab
Davide Maltoni
Vincenzo Lomonaco Lorenzo Pellegrini Gabriele Graffieti
Outline
1. Personal ResearchTrajectory andVision
2. Continual Learning: State-of-the-art
3. Rehearsal-free and Task-agnostic
Online Continual Learning
4. CurrentWork and Research Direction
PersonalResearch Trajectory
andVision
ResearchTrajectory andVision
I meet Davide Maltoni
who was working at
HTMs from 2011.
I read “On
Intelligence” and join
his quest for
understanding
intelligence and build it
in silicon.
MasterThesis Published:
“Comparing HTMs and CNNs
on Object RecognitionTasks”
2014
Visiting Scholar at Purdue
University.
Working on Continual
Reinforcement /
Unsupervised Learning.
Visiting Scholar at ENSTA
ParisTech.
Working on Continual for
Robotics and a more
comprehensive CL
framework definition.
2015 2017 2018
I defend my PhD
Dissertation “Continual
Learning with Deep
Architectures”.
Putting everything
together.
Post-Doc @ UniBo
on the same topic.
2019
We abandon HTM (1st Gen.) to
work on top of deep learning
directly with a focus on
Continual Learning.
In particular, on Continual
Learning from video sequences.
2016
Long-term vision: “Understand the key computational
principles of intelligence and build truly intelligent machines.”
Main research goal: “Closing the gap between the HTM
theory and current AI systems.”
OurWorks with HTMs (1st Gen.)
1. D. Maltoni, Pattern Recognition by HierarchicalTemporal
Memory,Technical Report, DEIS - University of Bologna technical
report, April 2011.
2. D. Maltoni and E.M. Rehn, Incremental Learning by Message
Passing in HierarchicalTemporal Memory in 5thWorkshop on
Artificial Neural Networks in Pattern Recognition (ANNPR12),
Trento (Italy), pp.24-35, September 2012.
3. E.M. Rehn and D. Maltoni, Incremental Learning by Message
Passing in HierarchicalTemporal Memory, Neural Computation,
vol.26, no.8, pp.1763-1809, August 2014.
4. D. Maltoni andV. Lomonaco, Semi-supervisedTuning from
Temporal Coherence, in International Conference on Pattern
Recognition (ICPR16), Cancun, Mexico, December 2016.
Semi-SupervisedTuning from
Temporal Coherence (with HTMs)
HTM theory
Principles of Intelligence
1. Hierarchical Learning
2. Sequence Learning
3. Continual Learning
4. Sparse Representations
5. Sensory-Motor Integration (Embodiment)
6. Distributed Parallel Modeling (Thousands BrainTheory)
7. … ?
Emerging Properties
Flexibility Robustness Scalability Efficiency Adaptation
Autonomy Generalization Compositionality Reasoning
Common Sense ...
Towards “Cortical Learning”
Neuroscience Grounding
PracticalFunctionality
Symbolic AI
Kernel Machines
Feed-Forward
NNs / LSTMs
CNNs
Deep-CNNs
Conv-LSTMs
Deep-RL
Continual
Learning
Cortical
Learning
HTM
(1st Gen.)
CLA
(2nd Gen.)
CLA
(3rd Gen.)
Other Approaches
HTM-based
Neural Networks Based
Bayesian Approaches
Analogism-based
Approaches Evolutionary Approaches
Continual Learning:
State-of-the-art
The Stability-Plasticity Dilemma
Stability-Plasticity Dilemma:
• Remember past concepts
• Learn new concepts
• Generalize
Biggest Problem in Deep Learning:
• Catastrophic Forgetting
The Stability-Plasticity Dilemma
Continual Learning: Approaches
T. Lesort,V. Lomonaco et al. Continual Learning for Robotics. pre-print arxiv arXiv:1907.00182 .
CL Framework
CL Algorithm
Mini-spot Robot from Boston Dynamics, 2018
T. Lesort,V. Lomonaco et al. Continual Learning for Robotics. pre-print arxiv arXiv:1907.00182 .
Rehearsal-free and Task-agnostic
Online Continual Learning
3 Short-term Research Objective for CL
1. Rehearsal-Free: Raw data cannot be stored and re-used
for rehearsal.
2. Task Agnostic: No use of supplementary task supervised
signal “t”.
3. Online: Bounded computational and memory
overheads, efficient, real-time updates (possibly one
data instance at a time).
T. Lesort,V. Lomonaco et al. Continual Learning for Robotics. pre-print arxiv arXiv:1907.00182 .
Task Agnostic Continual Learning
1. New Instances (NI)
2. New Classes (NC)
3. New Instances and Classes (NIC)
Initial Batch Incremental Batches
Τ
. . .
CORe50Website
Dataset, Benchmark, code and additional
information freely available at:
vlomonaco.github.io/core50
LomonacoV. and Maltoni D. CORe50: a New Dataset and Benchmark for Continuous Object Recognition. CoRL2017.
LomonacoV. and Maltoni D. CORe50: a New Dataset and Benchmark for Continuous Object Recognition. CoRL2017.
CORe50: aVideo Benchmark for CL
and Object Recognition/Detection
# Images 164,866
Format RGB-D
Image size 350x350
128x128
# Categories 10
# Obj. x Cat. 5
# Sessions 11
# img. x Sess. ~300
# Outdoor Sess. 3
Acquisition Sett. Hand held
LomonacoV. and Maltoni D. CORe50: a New Dataset and Benchmark for Continuous Object Recognition. CoRL2017.
CORe50: aVideo Benchmark for CL
and Object Recognition/Detection
Fine-Grained Continual Learning
LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
AR-1*
Rehearsal-free andTask Agnostic
Online Continual Learning
Maltoni D. and LomonacoV. Continuous Learning in Single-Incremental-Task Scenarios. Neural Networks Journal, 2019.
LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
AR-1*: Overview (with MobileNet-V1)
LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
AR-1*: Supervised / Unsupervised
Pre-Training Phase
● Supervised or
Unsupervised
Pre-Training from
ImageNet.
● Slowly Fine-tuned or
kept fixed.
● future direction:
unsupervised
co-training from
scratch.
LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
AR-1*: Regularization Phase
LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
AR-1*: Regularization Phase
● Computational
Efficient (independent
from the number of
training batches)
● Just one Fisher matrix
(running sum + max
clip)
● Importance of Batch
ReNormalization
LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
AR-1*: Architectural Phase
● CWR*: generalization of
CWR+ to handle
agnostically NI, NC and
NIC settings
● Dual-Memory system for
memory consolidation.
● Based on zero-init for new
classes, weights
consolidation and
finetuning for already
encountered classes.
LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
CORe50 - NICv2 Results
● (0%-92%) -45% avg. memory.
● (0%-94%) -49% avg. compute.
● -20% price in accuracy at
the end of last batch.
LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
CurrentWork and Research
Direction
Real-World Continual Learning on
Embedded Systems
Pellegrini L., Graffieti G. , LomonacoV. and Maltoni D. Towards Continual Learning on the Edge.To be published.
AR-1*: Closing the Accuracy Gap with
Latent Rehearsal
Pellegrini L., Graffieti G. , LomonacoV. and Maltoni D. Towards Continual Learning on the Edge.To be published.
AR-1*: Closing the Accuracy Gap with
Latent Rehearsal
Pellegrini L., Graffieti G. , LomonacoV. and Maltoni D. Towards Continual Learning on the Edge.To be published.
AR-1*: Sparse Representations
Pellegrini L., Graffieti G. , LomonacoV. and Maltoni D. Towards Continual Learning on the Edge.To be published.
● Imposing Sparsity of the
activactivation does not
affect accuracy from
~55% to ~35%.
● It has been shown that
sparsity may help the CL
process.
● Less memory overhead for
latent rehearsal.
FutureWorks and Research Direction
1. Latent Generative Replay
2. Lowering the amount of Supervision (Unsupervised
Reinforcement Learning, Active Learning)
3. Infer or make use of the sparse “task signal” (context
modulation)
4. Sequence Learning/ Temporal Coherence Integration
5. Improve robustness in real-world embedded
applications (Smartphone devices, Nao Robot, …)
Maltoni D. and LomonacoV. Semi-SupervisedTuning fromTemporal Coherence. ICPR 2016.
LomonacoV., Desai K., Maltoni D. and Culurciello, E. Continual Reinforcement Learning in 3D non-stationary
environments. preprint arxiv arXiv:1905.10112, 2019.
AR-1*: Closing the Accuracy Gap with
Latent Generative Replay
●
●
●
●
●
●
Pellegrini L., Graffieti G. , LomonacoV. and Maltoni D. Towards Continual Learning on the Edge.To be published.
Questions?
Introduction Meetup @ Numenta
16-09-2019
Vincenzo Lomonaco
vincenzo.lomonaco@unibo.it
Postdoctoral Researcher @ University of Bologna
Supervisor: Davide Maltoni

More Related Content

What's hot

Self-supervised Visual Learning 2020 - Xavier Giro-i-Nieto - UPC Barcelona
Self-supervised Visual Learning 2020 - Xavier Giro-i-Nieto - UPC BarcelonaSelf-supervised Visual Learning 2020 - Xavier Giro-i-Nieto - UPC Barcelona
Self-supervised Visual Learning 2020 - Xavier Giro-i-Nieto - UPC Barcelona
Universitat Politècnica de Catalunya
 
Ghent University and GUGC-K: Overview of Teaching and Research Activities
Ghent University and GUGC-K: Overview of Teaching and Research ActivitiesGhent University and GUGC-K: Overview of Teaching and Research Activities
Ghent University and GUGC-K: Overview of Teaching and Research Activities
Wesley De Neve
 
Kadir A_20160804_res_tea
Kadir A_20160804_res_teaKadir A_20160804_res_tea
Kadir A_20160804_res_teaKadir A Peker
 
Self-supervised Audiovisual Learning 2020 - Xavier Giro-i-Nieto - UPC Telecom...
Self-supervised Audiovisual Learning 2020 - Xavier Giro-i-Nieto - UPC Telecom...Self-supervised Audiovisual Learning 2020 - Xavier Giro-i-Nieto - UPC Telecom...
Self-supervised Audiovisual Learning 2020 - Xavier Giro-i-Nieto - UPC Telecom...
Universitat Politècnica de Catalunya
 
Deep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN Barcelona
Deep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN BarcelonaDeep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN Barcelona
Deep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN Barcelona
Universitat Politècnica de Catalunya
 
Inside the Black Box: How Does a Neural Network Understand Names? - Philip Bl...
Inside the Black Box: How Does a Neural Network Understand Names? - Philip Bl...Inside the Black Box: How Does a Neural Network Understand Names? - Philip Bl...
Inside the Black Box: How Does a Neural Network Understand Names? - Philip Bl...
Lucidworks
 
Self-Supervised Audio-Visual Learning - Xavier Giro - UPC TelecomBCN Barcelon...
Self-Supervised Audio-Visual Learning - Xavier Giro - UPC TelecomBCN Barcelon...Self-Supervised Audio-Visual Learning - Xavier Giro - UPC TelecomBCN Barcelon...
Self-Supervised Audio-Visual Learning - Xavier Giro - UPC TelecomBCN Barcelon...
Universitat Politècnica de Catalunya
 
Video Analysis with Convolutional Neural Networks (Master Computer Vision Bar...
Video Analysis with Convolutional Neural Networks (Master Computer Vision Bar...Video Analysis with Convolutional Neural Networks (Master Computer Vision Bar...
Video Analysis with Convolutional Neural Networks (Master Computer Vision Bar...
Universitat Politècnica de Catalunya
 
Deep Video Object Tracking - Xavier Giro - UPC Barcelona 2019
Deep Video Object Tracking - Xavier Giro - UPC Barcelona 2019Deep Video Object Tracking - Xavier Giro - UPC Barcelona 2019
Deep Video Object Tracking - Xavier Giro - UPC Barcelona 2019
Universitat Politècnica de Catalunya
 
COM623M1.doc.doc
COM623M1.doc.docCOM623M1.doc.doc
COM623M1.doc.docbutest
 
Digital logic introduction using fpg as resume
Digital logic introduction using fpg as resumeDigital logic introduction using fpg as resume
Digital logic introduction using fpg as resume
Alvi Milanisti
 
Disentangle motion, Foreground and Background Features in Videos
Disentangle motion, Foreground and Background Features in VideosDisentangle motion, Foreground and Background Features in Videos
Disentangle motion, Foreground and Background Features in Videos
Universitat Politècnica de Catalunya
 
Adelaide Jan 2009
Adelaide Jan 2009Adelaide Jan 2009
Adelaide Jan 2009
Jose Manuel Martins Ferreira
 
Deep Learning for Video: Action Recognition (UPC 2018)
Deep Learning for Video: Action Recognition (UPC 2018)Deep Learning for Video: Action Recognition (UPC 2018)
Deep Learning for Video: Action Recognition (UPC 2018)
Universitat Politècnica de Catalunya
 
Molecular autoencoder
Molecular autoencoderMolecular autoencoder
Molecular autoencoder
Dan Elton
 
Modeling and meta-modeling presentation at LTH, Sweden
Modeling and meta-modeling presentation at LTH, Sweden Modeling and meta-modeling presentation at LTH, Sweden
Modeling and meta-modeling presentation at LTH, Sweden Saïd Assar
 
Syllabus 6th sem UIET PU
Syllabus 6th sem UIET PUSyllabus 6th sem UIET PU
Syllabus 6th sem UIET PU
Shubham Bajaj
 

What's hot (17)

Self-supervised Visual Learning 2020 - Xavier Giro-i-Nieto - UPC Barcelona
Self-supervised Visual Learning 2020 - Xavier Giro-i-Nieto - UPC BarcelonaSelf-supervised Visual Learning 2020 - Xavier Giro-i-Nieto - UPC Barcelona
Self-supervised Visual Learning 2020 - Xavier Giro-i-Nieto - UPC Barcelona
 
Ghent University and GUGC-K: Overview of Teaching and Research Activities
Ghent University and GUGC-K: Overview of Teaching and Research ActivitiesGhent University and GUGC-K: Overview of Teaching and Research Activities
Ghent University and GUGC-K: Overview of Teaching and Research Activities
 
Kadir A_20160804_res_tea
Kadir A_20160804_res_teaKadir A_20160804_res_tea
Kadir A_20160804_res_tea
 
Self-supervised Audiovisual Learning 2020 - Xavier Giro-i-Nieto - UPC Telecom...
Self-supervised Audiovisual Learning 2020 - Xavier Giro-i-Nieto - UPC Telecom...Self-supervised Audiovisual Learning 2020 - Xavier Giro-i-Nieto - UPC Telecom...
Self-supervised Audiovisual Learning 2020 - Xavier Giro-i-Nieto - UPC Telecom...
 
Deep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN Barcelona
Deep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN BarcelonaDeep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN Barcelona
Deep Video Object Tracking 2020 - Xavier Giro - UPC TelecomBCN Barcelona
 
Inside the Black Box: How Does a Neural Network Understand Names? - Philip Bl...
Inside the Black Box: How Does a Neural Network Understand Names? - Philip Bl...Inside the Black Box: How Does a Neural Network Understand Names? - Philip Bl...
Inside the Black Box: How Does a Neural Network Understand Names? - Philip Bl...
 
Self-Supervised Audio-Visual Learning - Xavier Giro - UPC TelecomBCN Barcelon...
Self-Supervised Audio-Visual Learning - Xavier Giro - UPC TelecomBCN Barcelon...Self-Supervised Audio-Visual Learning - Xavier Giro - UPC TelecomBCN Barcelon...
Self-Supervised Audio-Visual Learning - Xavier Giro - UPC TelecomBCN Barcelon...
 
Video Analysis with Convolutional Neural Networks (Master Computer Vision Bar...
Video Analysis with Convolutional Neural Networks (Master Computer Vision Bar...Video Analysis with Convolutional Neural Networks (Master Computer Vision Bar...
Video Analysis with Convolutional Neural Networks (Master Computer Vision Bar...
 
Deep Video Object Tracking - Xavier Giro - UPC Barcelona 2019
Deep Video Object Tracking - Xavier Giro - UPC Barcelona 2019Deep Video Object Tracking - Xavier Giro - UPC Barcelona 2019
Deep Video Object Tracking - Xavier Giro - UPC Barcelona 2019
 
COM623M1.doc.doc
COM623M1.doc.docCOM623M1.doc.doc
COM623M1.doc.doc
 
Digital logic introduction using fpg as resume
Digital logic introduction using fpg as resumeDigital logic introduction using fpg as resume
Digital logic introduction using fpg as resume
 
Disentangle motion, Foreground and Background Features in Videos
Disentangle motion, Foreground and Background Features in VideosDisentangle motion, Foreground and Background Features in Videos
Disentangle motion, Foreground and Background Features in Videos
 
Adelaide Jan 2009
Adelaide Jan 2009Adelaide Jan 2009
Adelaide Jan 2009
 
Deep Learning for Video: Action Recognition (UPC 2018)
Deep Learning for Video: Action Recognition (UPC 2018)Deep Learning for Video: Action Recognition (UPC 2018)
Deep Learning for Video: Action Recognition (UPC 2018)
 
Molecular autoencoder
Molecular autoencoderMolecular autoencoder
Molecular autoencoder
 
Modeling and meta-modeling presentation at LTH, Sweden
Modeling and meta-modeling presentation at LTH, Sweden Modeling and meta-modeling presentation at LTH, Sweden
Modeling and meta-modeling presentation at LTH, Sweden
 
Syllabus 6th sem UIET PU
Syllabus 6th sem UIET PUSyllabus 6th sem UIET PU
Syllabus 6th sem UIET PU
 

Similar to Continual Learning: Another Step Towards Truly Intelligent Machines

Continual Learning with Deep Architectures Workshop @ Computer VISIONers Conf...
Continual Learning with Deep Architectures Workshop @ Computer VISIONers Conf...Continual Learning with Deep Architectures Workshop @ Computer VISIONers Conf...
Continual Learning with Deep Architectures Workshop @ Computer VISIONers Conf...
Vincenzo Lomonaco
 
Continual/Lifelong Learning with Deep Architectures, Vincenzo Lomonaco
Continual/Lifelong Learning with Deep Architectures, Vincenzo LomonacoContinual/Lifelong Learning with Deep Architectures, Vincenzo Lomonaco
Continual/Lifelong Learning with Deep Architectures, Vincenzo Lomonaco
Data Science Milan
 
Online Python Resources
Online Python ResourcesOnline Python Resources
Online Python Resources
Jonathan Fine
 
py04.pdf
py04.pdfpy04.pdf
py04.pdf
pradeepskvch
 
CORe50: a New Dataset and Benchmark for Continuous Object Recognition Poster
CORe50: a New Dataset and Benchmark for Continuous Object Recognition PosterCORe50: a New Dataset and Benchmark for Continuous Object Recognition Poster
CORe50: a New Dataset and Benchmark for Continuous Object Recognition Poster
Vincenzo Lomonaco
 
Deep Learning
Deep LearningDeep Learning
Deep Learning
Riri Kusumarani
 
Wageningen phenotype meeting
Wageningen phenotype meetingWageningen phenotype meeting
Wageningen phenotype meetingthehyve
 
Multimedia Information Retrieval: Bytes and pixels meet the challenges of hum...
Multimedia Information Retrieval: Bytes and pixels meet the challenges of hum...Multimedia Information Retrieval: Bytes and pixels meet the challenges of hum...
Multimedia Information Retrieval: Bytes and pixels meet the challenges of hum...
maranlar
 
Topologos
TopologosTopologos
Topologos
ESUG
 
Semantic, Cognitive and Perceptual Computing -Deep learning
Semantic, Cognitive and Perceptual Computing -Deep learning Semantic, Cognitive and Perceptual Computing -Deep learning
Semantic, Cognitive and Perceptual Computing -Deep learning
Artificial Intelligence Institute at UofSC
 
Deep Learning for Computer Vision: A comparision between Convolutional Neural...
Deep Learning for Computer Vision: A comparision between Convolutional Neural...Deep Learning for Computer Vision: A comparision between Convolutional Neural...
Deep Learning for Computer Vision: A comparision between Convolutional Neural...
Vincenzo Lomonaco
 
e-Sim approach at Universidade Aberta: presence, narrative and agency for imm...
e-Sim approach at Universidade Aberta: presence, narrative and agency for imm...e-Sim approach at Universidade Aberta: presence, narrative and agency for imm...
e-Sim approach at Universidade Aberta: presence, narrative and agency for imm...
Leonel Morgado
 
Technology and the Grand Challenge for Future Learning
Technology and the Grand Challenge for Future LearningTechnology and the Grand Challenge for Future Learning
Technology and the Grand Challenge for Future Learning
fridolin.wild
 
ESSLLI2016 DTS Lecture Day 5-1: Introduction to day 5
ESSLLI2016 DTS Lecture Day 5-1: Introduction to day 5ESSLLI2016 DTS Lecture Day 5-1: Introduction to day 5
ESSLLI2016 DTS Lecture Day 5-1: Introduction to day 5
Daisuke BEKKI
 
Herramientas 2.0 para el desarrollo de competencias profesionalizadoras
Herramientas 2.0 para el desarrollo de competencias profesionalizadorasHerramientas 2.0 para el desarrollo de competencias profesionalizadoras
Herramientas 2.0 para el desarrollo de competencias profesionalizadoras
Ismael Peña-López
 
Advances in Learning with Bayesian Networks - july 2015
Advances in Learning with Bayesian Networks - july 2015Advances in Learning with Bayesian Networks - july 2015
Advances in Learning with Bayesian Networks - july 2015
University of Nantes
 
final-day1-july2.pptx
final-day1-july2.pptxfinal-day1-july2.pptx
final-day1-july2.pptx
KartikGulati16
 
Hologram Lecturers and Tele-Presence Teachers in the Next Dimension
Hologram Lecturers and Tele-Presence Teachers in the Next DimensionHologram Lecturers and Tele-Presence Teachers in the Next Dimension
Hologram Lecturers and Tele-Presence Teachers in the Next Dimension
Zac Woolfitt
 

Similar to Continual Learning: Another Step Towards Truly Intelligent Machines (20)

Continual Learning with Deep Architectures Workshop @ Computer VISIONers Conf...
Continual Learning with Deep Architectures Workshop @ Computer VISIONers Conf...Continual Learning with Deep Architectures Workshop @ Computer VISIONers Conf...
Continual Learning with Deep Architectures Workshop @ Computer VISIONers Conf...
 
Continual/Lifelong Learning with Deep Architectures, Vincenzo Lomonaco
Continual/Lifelong Learning with Deep Architectures, Vincenzo LomonacoContinual/Lifelong Learning with Deep Architectures, Vincenzo Lomonaco
Continual/Lifelong Learning with Deep Architectures, Vincenzo Lomonaco
 
Online Python Resources
Online Python ResourcesOnline Python Resources
Online Python Resources
 
py04.pdf
py04.pdfpy04.pdf
py04.pdf
 
CORe50: a New Dataset and Benchmark for Continuous Object Recognition Poster
CORe50: a New Dataset and Benchmark for Continuous Object Recognition PosterCORe50: a New Dataset and Benchmark for Continuous Object Recognition Poster
CORe50: a New Dataset and Benchmark for Continuous Object Recognition Poster
 
Deep Learning
Deep LearningDeep Learning
Deep Learning
 
Wageningen phenotype meeting
Wageningen phenotype meetingWageningen phenotype meeting
Wageningen phenotype meeting
 
Multimedia Information Retrieval: Bytes and pixels meet the challenges of hum...
Multimedia Information Retrieval: Bytes and pixels meet the challenges of hum...Multimedia Information Retrieval: Bytes and pixels meet the challenges of hum...
Multimedia Information Retrieval: Bytes and pixels meet the challenges of hum...
 
Seyedi-curriculum vitae
Seyedi-curriculum vitaeSeyedi-curriculum vitae
Seyedi-curriculum vitae
 
Topologos
TopologosTopologos
Topologos
 
Semantic, Cognitive and Perceptual Computing -Deep learning
Semantic, Cognitive and Perceptual Computing -Deep learning Semantic, Cognitive and Perceptual Computing -Deep learning
Semantic, Cognitive and Perceptual Computing -Deep learning
 
Deep Learning for Computer Vision: A comparision between Convolutional Neural...
Deep Learning for Computer Vision: A comparision between Convolutional Neural...Deep Learning for Computer Vision: A comparision between Convolutional Neural...
Deep Learning for Computer Vision: A comparision between Convolutional Neural...
 
e-Sim approach at Universidade Aberta: presence, narrative and agency for imm...
e-Sim approach at Universidade Aberta: presence, narrative and agency for imm...e-Sim approach at Universidade Aberta: presence, narrative and agency for imm...
e-Sim approach at Universidade Aberta: presence, narrative and agency for imm...
 
Technology and the Grand Challenge for Future Learning
Technology and the Grand Challenge for Future LearningTechnology and the Grand Challenge for Future Learning
Technology and the Grand Challenge for Future Learning
 
ESSLLI2016 DTS Lecture Day 5-1: Introduction to day 5
ESSLLI2016 DTS Lecture Day 5-1: Introduction to day 5ESSLLI2016 DTS Lecture Day 5-1: Introduction to day 5
ESSLLI2016 DTS Lecture Day 5-1: Introduction to day 5
 
CV_academic
CV_academicCV_academic
CV_academic
 
Herramientas 2.0 para el desarrollo de competencias profesionalizadoras
Herramientas 2.0 para el desarrollo de competencias profesionalizadorasHerramientas 2.0 para el desarrollo de competencias profesionalizadoras
Herramientas 2.0 para el desarrollo de competencias profesionalizadoras
 
Advances in Learning with Bayesian Networks - july 2015
Advances in Learning with Bayesian Networks - july 2015Advances in Learning with Bayesian Networks - july 2015
Advances in Learning with Bayesian Networks - july 2015
 
final-day1-july2.pptx
final-day1-july2.pptxfinal-day1-july2.pptx
final-day1-july2.pptx
 
Hologram Lecturers and Tele-Presence Teachers in the Next Dimension
Hologram Lecturers and Tele-Presence Teachers in the Next DimensionHologram Lecturers and Tele-Presence Teachers in the Next Dimension
Hologram Lecturers and Tele-Presence Teachers in the Next Dimension
 

More from Vincenzo Lomonaco

2023-08-22 CoLLAs Tutorial - Beyond CIL.pdf
2023-08-22 CoLLAs Tutorial - Beyond CIL.pdf2023-08-22 CoLLAs Tutorial - Beyond CIL.pdf
2023-08-22 CoLLAs Tutorial - Beyond CIL.pdf
Vincenzo Lomonaco
 
Continual Learning with Deep Architectures - Tutorial ICML 2021
Continual Learning with Deep Architectures - Tutorial ICML 2021Continual Learning with Deep Architectures - Tutorial ICML 2021
Continual Learning with Deep Architectures - Tutorial ICML 2021
Vincenzo Lomonaco
 
Continual Learning for Robotics
Continual Learning for RoboticsContinual Learning for Robotics
Continual Learning for Robotics
Vincenzo Lomonaco
 
Don't forget, there is more than forgetting: new metrics for Continual Learni...
Don't forget, there is more than forgetting: new metrics for Continual Learni...Don't forget, there is more than forgetting: new metrics for Continual Learni...
Don't forget, there is more than forgetting: new metrics for Continual Learni...
Vincenzo Lomonaco
 
Open-Source Frameworks for Deep Learning: an Overview
Open-Source Frameworks for Deep Learning: an OverviewOpen-Source Frameworks for Deep Learning: an Overview
Open-Source Frameworks for Deep Learning: an Overview
Vincenzo Lomonaco
 
CORe50: a New Dataset and Benchmark for Continual Learning and Object Recogni...
CORe50: a New Dataset and Benchmark for Continual Learning and Object Recogni...CORe50: a New Dataset and Benchmark for Continual Learning and Object Recogni...
CORe50: a New Dataset and Benchmark for Continual Learning and Object Recogni...
Vincenzo Lomonaco
 
Continuous Learning with Deep Architectures
Continuous Learning with Deep ArchitecturesContinuous Learning with Deep Architectures
Continuous Learning with Deep Architectures
Vincenzo Lomonaco
 
Continuous Unsupervised Training of Deep Architectures
Continuous Unsupervised Training of Deep ArchitecturesContinuous Unsupervised Training of Deep Architectures
Continuous Unsupervised Training of Deep Architectures
Vincenzo Lomonaco
 
Comparing Incremental Learning Strategies for Convolutional Neural Networks
Comparing Incremental Learning Strategies for Convolutional Neural NetworksComparing Incremental Learning Strategies for Convolutional Neural Networks
Comparing Incremental Learning Strategies for Convolutional Neural Networks
Vincenzo Lomonaco
 
Deep Learning for Computer Vision: A comparision between Convolutional Neural...
Deep Learning for Computer Vision: A comparision between Convolutional Neural...Deep Learning for Computer Vision: A comparision between Convolutional Neural...
Deep Learning for Computer Vision: A comparision between Convolutional Neural...
Vincenzo Lomonaco
 
A Framework for Deadlock Detection in Java
A Framework for Deadlock Detection in JavaA Framework for Deadlock Detection in Java
A Framework for Deadlock Detection in Java
Vincenzo Lomonaco
 
Deep Learning libraries and first experiments with Theano
Deep Learning libraries and first experiments with TheanoDeep Learning libraries and first experiments with Theano
Deep Learning libraries and first experiments with Theano
Vincenzo Lomonaco
 
Word2vec on the italian language: first experiments
Word2vec on the italian language: first experimentsWord2vec on the italian language: first experiments
Word2vec on the italian language: first experiments
Vincenzo Lomonaco
 
Machine Learning for Automated Reasoning: An Overview
Machine Learning for Automated Reasoning: An OverviewMachine Learning for Automated Reasoning: An Overview
Machine Learning for Automated Reasoning: An Overview
Vincenzo Lomonaco
 

More from Vincenzo Lomonaco (14)

2023-08-22 CoLLAs Tutorial - Beyond CIL.pdf
2023-08-22 CoLLAs Tutorial - Beyond CIL.pdf2023-08-22 CoLLAs Tutorial - Beyond CIL.pdf
2023-08-22 CoLLAs Tutorial - Beyond CIL.pdf
 
Continual Learning with Deep Architectures - Tutorial ICML 2021
Continual Learning with Deep Architectures - Tutorial ICML 2021Continual Learning with Deep Architectures - Tutorial ICML 2021
Continual Learning with Deep Architectures - Tutorial ICML 2021
 
Continual Learning for Robotics
Continual Learning for RoboticsContinual Learning for Robotics
Continual Learning for Robotics
 
Don't forget, there is more than forgetting: new metrics for Continual Learni...
Don't forget, there is more than forgetting: new metrics for Continual Learni...Don't forget, there is more than forgetting: new metrics for Continual Learni...
Don't forget, there is more than forgetting: new metrics for Continual Learni...
 
Open-Source Frameworks for Deep Learning: an Overview
Open-Source Frameworks for Deep Learning: an OverviewOpen-Source Frameworks for Deep Learning: an Overview
Open-Source Frameworks for Deep Learning: an Overview
 
CORe50: a New Dataset and Benchmark for Continual Learning and Object Recogni...
CORe50: a New Dataset and Benchmark for Continual Learning and Object Recogni...CORe50: a New Dataset and Benchmark for Continual Learning and Object Recogni...
CORe50: a New Dataset and Benchmark for Continual Learning and Object Recogni...
 
Continuous Learning with Deep Architectures
Continuous Learning with Deep ArchitecturesContinuous Learning with Deep Architectures
Continuous Learning with Deep Architectures
 
Continuous Unsupervised Training of Deep Architectures
Continuous Unsupervised Training of Deep ArchitecturesContinuous Unsupervised Training of Deep Architectures
Continuous Unsupervised Training of Deep Architectures
 
Comparing Incremental Learning Strategies for Convolutional Neural Networks
Comparing Incremental Learning Strategies for Convolutional Neural NetworksComparing Incremental Learning Strategies for Convolutional Neural Networks
Comparing Incremental Learning Strategies for Convolutional Neural Networks
 
Deep Learning for Computer Vision: A comparision between Convolutional Neural...
Deep Learning for Computer Vision: A comparision between Convolutional Neural...Deep Learning for Computer Vision: A comparision between Convolutional Neural...
Deep Learning for Computer Vision: A comparision between Convolutional Neural...
 
A Framework for Deadlock Detection in Java
A Framework for Deadlock Detection in JavaA Framework for Deadlock Detection in Java
A Framework for Deadlock Detection in Java
 
Deep Learning libraries and first experiments with Theano
Deep Learning libraries and first experiments with TheanoDeep Learning libraries and first experiments with Theano
Deep Learning libraries and first experiments with Theano
 
Word2vec on the italian language: first experiments
Word2vec on the italian language: first experimentsWord2vec on the italian language: first experiments
Word2vec on the italian language: first experiments
 
Machine Learning for Automated Reasoning: An Overview
Machine Learning for Automated Reasoning: An OverviewMachine Learning for Automated Reasoning: An Overview
Machine Learning for Automated Reasoning: An Overview
 

Recently uploaded

FAIR & AI Ready KGs for Explainable Predictions
FAIR & AI Ready KGs for Explainable PredictionsFAIR & AI Ready KGs for Explainable Predictions
FAIR & AI Ready KGs for Explainable Predictions
Michel Dumontier
 
GBSN - Biochemistry (Unit 5) Chemistry of Lipids
GBSN - Biochemistry (Unit 5) Chemistry of LipidsGBSN - Biochemistry (Unit 5) Chemistry of Lipids
GBSN - Biochemistry (Unit 5) Chemistry of Lipids
Areesha Ahmad
 
extra-chromosomal-inheritance[1].pptx.pdfpdf
extra-chromosomal-inheritance[1].pptx.pdfpdfextra-chromosomal-inheritance[1].pptx.pdfpdf
extra-chromosomal-inheritance[1].pptx.pdfpdf
DiyaBiswas10
 
plant biotechnology Lecture note ppt.pptx
plant biotechnology Lecture note ppt.pptxplant biotechnology Lecture note ppt.pptx
plant biotechnology Lecture note ppt.pptx
yusufzako14
 
Comparative structure of adrenal gland in vertebrates
Comparative structure of adrenal gland in vertebratesComparative structure of adrenal gland in vertebrates
Comparative structure of adrenal gland in vertebrates
sachin783648
 
Anemia_ different types_causes_ conditions
Anemia_ different types_causes_ conditionsAnemia_ different types_causes_ conditions
Anemia_ different types_causes_ conditions
muralinath2
 
NuGOweek 2024 Ghent - programme - final version
NuGOweek 2024 Ghent - programme - final versionNuGOweek 2024 Ghent - programme - final version
NuGOweek 2024 Ghent - programme - final version
pablovgd
 
Mammalian Pineal Body Structure and Also Functions
Mammalian Pineal Body Structure and Also FunctionsMammalian Pineal Body Structure and Also Functions
Mammalian Pineal Body Structure and Also Functions
YOGESH DOGRA
 
insect taxonomy importance systematics and classification
insect taxonomy importance systematics and classificationinsect taxonomy importance systematics and classification
insect taxonomy importance systematics and classification
anitaento25
 
Structural Classification Of Protein (SCOP)
Structural Classification Of Protein  (SCOP)Structural Classification Of Protein  (SCOP)
Structural Classification Of Protein (SCOP)
aishnasrivastava
 
Penicillin...........................pptx
Penicillin...........................pptxPenicillin...........................pptx
Penicillin...........................pptx
Cherry
 
Orion Air Quality Monitoring Systems - CWS
Orion Air Quality Monitoring Systems - CWSOrion Air Quality Monitoring Systems - CWS
Orion Air Quality Monitoring Systems - CWS
Columbia Weather Systems
 
(May 29th, 2024) Advancements in Intravital Microscopy- Insights for Preclini...
(May 29th, 2024) Advancements in Intravital Microscopy- Insights for Preclini...(May 29th, 2024) Advancements in Intravital Microscopy- Insights for Preclini...
(May 29th, 2024) Advancements in Intravital Microscopy- Insights for Preclini...
Scintica Instrumentation
 
THE IMPORTANCE OF MARTIAN ATMOSPHERE SAMPLE RETURN.
THE IMPORTANCE OF MARTIAN ATMOSPHERE SAMPLE RETURN.THE IMPORTANCE OF MARTIAN ATMOSPHERE SAMPLE RETURN.
THE IMPORTANCE OF MARTIAN ATMOSPHERE SAMPLE RETURN.
Sérgio Sacani
 
Structures and textures of metamorphic rocks
Structures and textures of metamorphic rocksStructures and textures of metamorphic rocks
Structures and textures of metamorphic rocks
kumarmathi863
 
justice-and-fairness-ethics with example
justice-and-fairness-ethics with examplejustice-and-fairness-ethics with example
justice-and-fairness-ethics with example
azzyixes
 
Richard's aventures in two entangled wonderlands
Richard's aventures in two entangled wonderlandsRichard's aventures in two entangled wonderlands
Richard's aventures in two entangled wonderlands
Richard Gill
 
filosofia boliviana introducción jsjdjd.pptx
filosofia boliviana introducción jsjdjd.pptxfilosofia boliviana introducción jsjdjd.pptx
filosofia boliviana introducción jsjdjd.pptx
IvanMallco1
 
general properties of oerganologametal.ppt
general properties of oerganologametal.pptgeneral properties of oerganologametal.ppt
general properties of oerganologametal.ppt
IqrimaNabilatulhusni
 
Lab report on liquid viscosity of glycerin
Lab report on liquid viscosity of glycerinLab report on liquid viscosity of glycerin
Lab report on liquid viscosity of glycerin
ossaicprecious19
 

Recently uploaded (20)

FAIR & AI Ready KGs for Explainable Predictions
FAIR & AI Ready KGs for Explainable PredictionsFAIR & AI Ready KGs for Explainable Predictions
FAIR & AI Ready KGs for Explainable Predictions
 
GBSN - Biochemistry (Unit 5) Chemistry of Lipids
GBSN - Biochemistry (Unit 5) Chemistry of LipidsGBSN - Biochemistry (Unit 5) Chemistry of Lipids
GBSN - Biochemistry (Unit 5) Chemistry of Lipids
 
extra-chromosomal-inheritance[1].pptx.pdfpdf
extra-chromosomal-inheritance[1].pptx.pdfpdfextra-chromosomal-inheritance[1].pptx.pdfpdf
extra-chromosomal-inheritance[1].pptx.pdfpdf
 
plant biotechnology Lecture note ppt.pptx
plant biotechnology Lecture note ppt.pptxplant biotechnology Lecture note ppt.pptx
plant biotechnology Lecture note ppt.pptx
 
Comparative structure of adrenal gland in vertebrates
Comparative structure of adrenal gland in vertebratesComparative structure of adrenal gland in vertebrates
Comparative structure of adrenal gland in vertebrates
 
Anemia_ different types_causes_ conditions
Anemia_ different types_causes_ conditionsAnemia_ different types_causes_ conditions
Anemia_ different types_causes_ conditions
 
NuGOweek 2024 Ghent - programme - final version
NuGOweek 2024 Ghent - programme - final versionNuGOweek 2024 Ghent - programme - final version
NuGOweek 2024 Ghent - programme - final version
 
Mammalian Pineal Body Structure and Also Functions
Mammalian Pineal Body Structure and Also FunctionsMammalian Pineal Body Structure and Also Functions
Mammalian Pineal Body Structure and Also Functions
 
insect taxonomy importance systematics and classification
insect taxonomy importance systematics and classificationinsect taxonomy importance systematics and classification
insect taxonomy importance systematics and classification
 
Structural Classification Of Protein (SCOP)
Structural Classification Of Protein  (SCOP)Structural Classification Of Protein  (SCOP)
Structural Classification Of Protein (SCOP)
 
Penicillin...........................pptx
Penicillin...........................pptxPenicillin...........................pptx
Penicillin...........................pptx
 
Orion Air Quality Monitoring Systems - CWS
Orion Air Quality Monitoring Systems - CWSOrion Air Quality Monitoring Systems - CWS
Orion Air Quality Monitoring Systems - CWS
 
(May 29th, 2024) Advancements in Intravital Microscopy- Insights for Preclini...
(May 29th, 2024) Advancements in Intravital Microscopy- Insights for Preclini...(May 29th, 2024) Advancements in Intravital Microscopy- Insights for Preclini...
(May 29th, 2024) Advancements in Intravital Microscopy- Insights for Preclini...
 
THE IMPORTANCE OF MARTIAN ATMOSPHERE SAMPLE RETURN.
THE IMPORTANCE OF MARTIAN ATMOSPHERE SAMPLE RETURN.THE IMPORTANCE OF MARTIAN ATMOSPHERE SAMPLE RETURN.
THE IMPORTANCE OF MARTIAN ATMOSPHERE SAMPLE RETURN.
 
Structures and textures of metamorphic rocks
Structures and textures of metamorphic rocksStructures and textures of metamorphic rocks
Structures and textures of metamorphic rocks
 
justice-and-fairness-ethics with example
justice-and-fairness-ethics with examplejustice-and-fairness-ethics with example
justice-and-fairness-ethics with example
 
Richard's aventures in two entangled wonderlands
Richard's aventures in two entangled wonderlandsRichard's aventures in two entangled wonderlands
Richard's aventures in two entangled wonderlands
 
filosofia boliviana introducción jsjdjd.pptx
filosofia boliviana introducción jsjdjd.pptxfilosofia boliviana introducción jsjdjd.pptx
filosofia boliviana introducción jsjdjd.pptx
 
general properties of oerganologametal.ppt
general properties of oerganologametal.pptgeneral properties of oerganologametal.ppt
general properties of oerganologametal.ppt
 
Lab report on liquid viscosity of glycerin
Lab report on liquid viscosity of glycerinLab report on liquid viscosity of glycerin
Lab report on liquid viscosity of glycerin
 

Continual Learning: Another Step Towards Truly Intelligent Machines

  • 1. Continual Learning: Another Step TowardsTruly Intelligent Machines Introduction Meetup @ Numenta 16-09-2019 Vincenzo Lomonaco vincenzo.lomonaco@unibo.it Postdoctoral Researcher @ University of Bologna Supervisor: Davide Maltoni
  • 2. About me • Post-Doc @ University of Bologna • Research Affiliate @ AI Labs • Teaching Assistant of the courses Machine Learning and Computer Architectures @ UniBo • Author andTechnical reviewer of the online course Deep Learning with R and book R Deep Learning Essentials. • Co-Founder and President of ContinualAI.org • Co-Founder and Board Member of Data Science Bologna and AIforPeople.org
  • 3. What’s ContinualAI? • ContinualAI is a non-profit research organization and the largest research community on Continual Learning for AI. • It counts more than 550+ members in 17 different time-zones and from top-notch research institutions. • Learn more about ContinualAI at www.continualai.org
  • 5. Machine Intelligence @ BioLab Davide Maltoni Vincenzo Lomonaco Lorenzo Pellegrini Gabriele Graffieti
  • 6. Outline 1. Personal ResearchTrajectory andVision 2. Continual Learning: State-of-the-art 3. Rehearsal-free and Task-agnostic Online Continual Learning 4. CurrentWork and Research Direction
  • 8. ResearchTrajectory andVision I meet Davide Maltoni who was working at HTMs from 2011. I read “On Intelligence” and join his quest for understanding intelligence and build it in silicon. MasterThesis Published: “Comparing HTMs and CNNs on Object RecognitionTasks” 2014 Visiting Scholar at Purdue University. Working on Continual Reinforcement / Unsupervised Learning. Visiting Scholar at ENSTA ParisTech. Working on Continual for Robotics and a more comprehensive CL framework definition. 2015 2017 2018 I defend my PhD Dissertation “Continual Learning with Deep Architectures”. Putting everything together. Post-Doc @ UniBo on the same topic. 2019 We abandon HTM (1st Gen.) to work on top of deep learning directly with a focus on Continual Learning. In particular, on Continual Learning from video sequences. 2016 Long-term vision: “Understand the key computational principles of intelligence and build truly intelligent machines.” Main research goal: “Closing the gap between the HTM theory and current AI systems.”
  • 9. OurWorks with HTMs (1st Gen.) 1. D. Maltoni, Pattern Recognition by HierarchicalTemporal Memory,Technical Report, DEIS - University of Bologna technical report, April 2011. 2. D. Maltoni and E.M. Rehn, Incremental Learning by Message Passing in HierarchicalTemporal Memory in 5thWorkshop on Artificial Neural Networks in Pattern Recognition (ANNPR12), Trento (Italy), pp.24-35, September 2012. 3. E.M. Rehn and D. Maltoni, Incremental Learning by Message Passing in HierarchicalTemporal Memory, Neural Computation, vol.26, no.8, pp.1763-1809, August 2014. 4. D. Maltoni andV. Lomonaco, Semi-supervisedTuning from Temporal Coherence, in International Conference on Pattern Recognition (ICPR16), Cancun, Mexico, December 2016.
  • 11. HTM theory Principles of Intelligence 1. Hierarchical Learning 2. Sequence Learning 3. Continual Learning 4. Sparse Representations 5. Sensory-Motor Integration (Embodiment) 6. Distributed Parallel Modeling (Thousands BrainTheory) 7. … ? Emerging Properties Flexibility Robustness Scalability Efficiency Adaptation Autonomy Generalization Compositionality Reasoning Common Sense ...
  • 12. Towards “Cortical Learning” Neuroscience Grounding PracticalFunctionality Symbolic AI Kernel Machines Feed-Forward NNs / LSTMs CNNs Deep-CNNs Conv-LSTMs Deep-RL Continual Learning Cortical Learning HTM (1st Gen.) CLA (2nd Gen.) CLA (3rd Gen.) Other Approaches HTM-based Neural Networks Based Bayesian Approaches Analogism-based Approaches Evolutionary Approaches
  • 14. The Stability-Plasticity Dilemma Stability-Plasticity Dilemma: • Remember past concepts • Learn new concepts • Generalize Biggest Problem in Deep Learning: • Catastrophic Forgetting
  • 16. Continual Learning: Approaches T. Lesort,V. Lomonaco et al. Continual Learning for Robotics. pre-print arxiv arXiv:1907.00182 .
  • 17. CL Framework CL Algorithm Mini-spot Robot from Boston Dynamics, 2018 T. Lesort,V. Lomonaco et al. Continual Learning for Robotics. pre-print arxiv arXiv:1907.00182 .
  • 19. 3 Short-term Research Objective for CL 1. Rehearsal-Free: Raw data cannot be stored and re-used for rehearsal. 2. Task Agnostic: No use of supplementary task supervised signal “t”. 3. Online: Bounded computational and memory overheads, efficient, real-time updates (possibly one data instance at a time). T. Lesort,V. Lomonaco et al. Continual Learning for Robotics. pre-print arxiv arXiv:1907.00182 .
  • 20. Task Agnostic Continual Learning 1. New Instances (NI) 2. New Classes (NC) 3. New Instances and Classes (NIC) Initial Batch Incremental Batches Τ . . .
  • 21. CORe50Website Dataset, Benchmark, code and additional information freely available at: vlomonaco.github.io/core50 LomonacoV. and Maltoni D. CORe50: a New Dataset and Benchmark for Continuous Object Recognition. CoRL2017.
  • 22. LomonacoV. and Maltoni D. CORe50: a New Dataset and Benchmark for Continuous Object Recognition. CoRL2017. CORe50: aVideo Benchmark for CL and Object Recognition/Detection
  • 23. # Images 164,866 Format RGB-D Image size 350x350 128x128 # Categories 10 # Obj. x Cat. 5 # Sessions 11 # img. x Sess. ~300 # Outdoor Sess. 3 Acquisition Sett. Hand held LomonacoV. and Maltoni D. CORe50: a New Dataset and Benchmark for Continuous Object Recognition. CoRL2017. CORe50: aVideo Benchmark for CL and Object Recognition/Detection
  • 24. Fine-Grained Continual Learning LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
  • 25. AR-1* Rehearsal-free andTask Agnostic Online Continual Learning Maltoni D. and LomonacoV. Continuous Learning in Single-Incremental-Task Scenarios. Neural Networks Journal, 2019. LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
  • 26. AR-1*: Overview (with MobileNet-V1) LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
  • 27. AR-1*: Supervised / Unsupervised Pre-Training Phase ● Supervised or Unsupervised Pre-Training from ImageNet. ● Slowly Fine-tuned or kept fixed. ● future direction: unsupervised co-training from scratch. LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
  • 28. AR-1*: Regularization Phase LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
  • 29. AR-1*: Regularization Phase ● Computational Efficient (independent from the number of training batches) ● Just one Fisher matrix (running sum + max clip) ● Importance of Batch ReNormalization LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
  • 30. AR-1*: Architectural Phase ● CWR*: generalization of CWR+ to handle agnostically NI, NC and NIC settings ● Dual-Memory system for memory consolidation. ● Based on zero-init for new classes, weights consolidation and finetuning for already encountered classes. LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
  • 31. CORe50 - NICv2 Results ● (0%-92%) -45% avg. memory. ● (0%-94%) -49% avg. compute. ● -20% price in accuracy at the end of last batch. LomonacoV., Maltoni D., Pellegrini L. Fine-Grained Continual Learning. Preprint arxiv arXiv:1907.03799, 2019.
  • 33. Real-World Continual Learning on Embedded Systems Pellegrini L., Graffieti G. , LomonacoV. and Maltoni D. Towards Continual Learning on the Edge.To be published.
  • 34. AR-1*: Closing the Accuracy Gap with Latent Rehearsal Pellegrini L., Graffieti G. , LomonacoV. and Maltoni D. Towards Continual Learning on the Edge.To be published.
  • 35. AR-1*: Closing the Accuracy Gap with Latent Rehearsal Pellegrini L., Graffieti G. , LomonacoV. and Maltoni D. Towards Continual Learning on the Edge.To be published.
  • 36. AR-1*: Sparse Representations Pellegrini L., Graffieti G. , LomonacoV. and Maltoni D. Towards Continual Learning on the Edge.To be published. ● Imposing Sparsity of the activactivation does not affect accuracy from ~55% to ~35%. ● It has been shown that sparsity may help the CL process. ● Less memory overhead for latent rehearsal.
  • 37. FutureWorks and Research Direction 1. Latent Generative Replay 2. Lowering the amount of Supervision (Unsupervised Reinforcement Learning, Active Learning) 3. Infer or make use of the sparse “task signal” (context modulation) 4. Sequence Learning/ Temporal Coherence Integration 5. Improve robustness in real-world embedded applications (Smartphone devices, Nao Robot, …) Maltoni D. and LomonacoV. Semi-SupervisedTuning fromTemporal Coherence. ICPR 2016. LomonacoV., Desai K., Maltoni D. and Culurciello, E. Continual Reinforcement Learning in 3D non-stationary environments. preprint arxiv arXiv:1905.10112, 2019.
  • 38. AR-1*: Closing the Accuracy Gap with Latent Generative Replay ● ● ● ● ● ● Pellegrini L., Graffieti G. , LomonacoV. and Maltoni D. Towards Continual Learning on the Edge.To be published.
  • 39. Questions? Introduction Meetup @ Numenta 16-09-2019 Vincenzo Lomonaco vincenzo.lomonaco@unibo.it Postdoctoral Researcher @ University of Bologna Supervisor: Davide Maltoni