Published Sep 14, 2020

SWaV: Unsupervised Learning of Visual Features by Contrasting Cluster Assignments (Mathilde Caron)

Explore the innovative SWaV algorithm with insights from Facebook Research's Mathilde Caron as it transforms unsupervised learning through advanced clustering techniques and multi-crop data augmentation, significantly enhancing image recognition in data-scarce environments.

Episode Highlights

Topics covered

Episode Highlights

Related Episodes

#55 Self-Supervised Vision Models (Dr. Ishan Misra - FAIR).
Answers 383 questions
CURL: Contrastive Unsupervised Representations for Reinforcement Learning
Answers 383 questions
#52 - Unadversarial Examples (Hadi Salman, MIT)
Answers 383 questions
#86 - Prof. YANN LECUN and Dr. RANDALL BALESTRIERO - SSL, Data Augmentation, Reward isn't enough [NEURIPS2022]
Answers 383 questions
Facebook Research - Unsupervised Translation of Programming Languages
Answers 383 questions
#68 DR. WALID SABA 2.0 - Natural Language Understanding [UNPLUGGED]
Answers 383 questions
Dr. Paul Lessard - Categorical/Structured Deep Learning
Answers 383 questions
WelcomeAIOverlords (Zak Jost)
Answers 383 questions
Sayak Paul
Answers 383 questions
#040 - Adversarial Examples (Dr. Nicholas Carlini, Dr. Wieland Brendel, Florian Tramèr)
Answers 383 questions
#046 The Great ML Stagnation (Mark Saroufim and Dr. Mathew Salvaris)
Answers 383 questions
#032- Simon Kornblith / GoogleAI - SimCLR and Paper Haul!
Answers 383 questions
One Shot and Metric Learning - Quadruplet Loss (Machine Learning Dojo)
Answers 383 questions
061: Interpolation, Extrapolation and Linearisation (Prof. Yann LeCun, Dr. Randall Balestriero)
Answers 383 questions
Kernels!
Answers 383 questions

Dexa/Machine Learning Street Talk (MLST)

SWaV: Unsupervised Learning of Visual Features by Contrasting Cluster Assignments (Mathilde Caron)

Topics covered

Popular Clips

Unsupervised Learning Insights

Clustering and Attention

Unsupervised Learning Insights

Unsupervised Feature Learning

Embeddings and Prototypes

Clustering with Constraints

Unsupervised Learning Insights

Enhancing Clustering Stability

Self-Supervised Learning

Clustering in Contrastive Learning

Overcoming Research Hurdles

Unsupervised Data Augmentation

Self-Supervised Learning Insights

Unsupervised Learning Insights

Episode Highlights

Clustering Approaches

Data Augmentation Strategies

Self-Supervised Learning

Related Episodes

#55 Self-Supervised Vision Models (Dr. Ishan Misra - FAIR).

CURL: Contrastive Unsupervised Representations for Reinforcement Learning

#52 - Unadversarial Examples (Hadi Salman, MIT)

#86 - Prof. YANN LECUN and Dr. RANDALL BALESTRIERO - SSL, Data Augmentation, Reward isn't enough [NEURIPS2022]

Facebook Research - Unsupervised Translation of Programming Languages

#68 DR. WALID SABA 2.0 - Natural Language Understanding [UNPLUGGED]

Dr. Paul Lessard - Categorical/Structured Deep Learning

WelcomeAIOverlords (Zak Jost)

Sayak Paul

#040 - Adversarial Examples (Dr. Nicholas Carlini, Dr. Wieland Brendel, Florian Tramèr)

#046 The Great ML Stagnation (Mark Saroufim and Dr. Mathew Salvaris)

#032- Simon Kornblith / GoogleAI - SimCLR and Paper Haul!

One Shot and Metric Learning - Quadruplet Loss (Machine Learning Dojo)

061: Interpolation, Extrapolation and Linearisation (Prof. Yann LeCun, Dr. Randall Balestriero)

Kernels!

SWaV: Unsupervised Learning of Visual Features by Contrasting Cluster Assignments (Mathilde Caron)

Topics covered

Popular Clips

Episode Highlights

Clustering ApproachesThe SWaV algorithm revolutionizes unsupervised learning by leveraging clustering to efficiently learn visual features. It addresses computational challenges of traditional methods, offering a more stable and effective approach to image representation.

Clustering Approaches

Data Augmentation Strategies

Self-Supervised Learning

Related Episodes