Open in app

Sign In

Write

Sign In

Alex Adam
Alex Adam

91 Followers

Home

About

Published in Towards Data Science

·Mar 12, 2021

Using CNN Extracted Features In Fun New Ways

Segmenting images using classification models. — Unpacking the features learned by a deep convolutional neural network (CNN) is a dauting task. Going through each layer to either visualize filters or features scales poorly with network depth, and although some cool figures can be created with this process, the result can be quite abstract, even psychedelic. Techniques…

Convolutional Network

4 min read

Visualizing Image Similarities
Visualizing Image Similarities
Convolutional Network

4 min read


Published in Towards Data Science

·May 13, 2020

Ensemble Robustness to Adversarial Examples

Last summer I had the pleasure of working with a talented undergraduate researcher named Romain Speciel on a project that looked at how to regularize model ensembles in a way that improves robustness to adversarial examples. Our main concern was reducing the transferability of adversarial examples between models as that…

Machine Learning

7 min read

Ensemble Robustness to Adversarial Examples
Ensemble Robustness to Adversarial Examples
Machine Learning

7 min read


Published in Towards Data Science

·Apr 25, 2020

Increasing Interpretability to Improve Model Robustness

A recent attempt to improve the robustness of convolutional neural networks (CNNs) on image classification tasks has revealed an interesting link between robustness and interpretability. Models trained using adversarial training, a training procedure that augments training data with adversarial examples, have input gradients that qualitatively appear to be using more…

Deep Learning

6 min read

Increasing Interpretability to Improve Model Robustness
Increasing Interpretability to Improve Model Robustness
Deep Learning

6 min read


Published in Towards Data Science

·Sep 14, 2019

Neural Architecture Search — Limitations and Extensions

For the past couple of years, researchers and companies have been trying to make deep learning more accessible to non-experts by providing access to pre-trained computer vision or machine translation models. Using a pre-trained model for another task is known as transfer learning, but it still requires sufficient expertise to…

Machine Learning

11 min read

Neural Architecture Search — Limitations and Extensions
Neural Architecture Search — Limitations and Extensions
Machine Learning

11 min read


Published in Towards Data Science

·Sep 8, 2019

Adversarial Examples — Rethinking the Definition

Adversarial examples are a large obstacle for a variety of machine learning systems to overcome. Their existence shows the tendency of models to rely on unreliable features to maximize performance, which if perturbed, can cause misclassifications with potentially catastrophic consequences. The informal definition of an adversarial example is an input…

Machine Learning

8 min read

Adversarial Examples — Rethinking the Definition
Adversarial Examples — Rethinking the Definition
Machine Learning

8 min read

Alex Adam

Alex Adam

91 Followers

PhD Student in Machine Learning

Following
  • Jonathan Hui

    Jonathan Hui

  • Felix Laumann

    Felix Laumann

  • Ceshine Lee

    Ceshine Lee

  • Neuromation

    Neuromation

  • CBMP

    CBMP

Help

Status

Writers

Blog

Careers

Privacy

Terms

About

Text to speech