Khurram Javed

kjaved (at) ualberta (dot) ca

I am a M.Sc student at the RLAI lab working with Martha White. I'm interested in building computational intelligence that learns online. Currently, I'm working on scalable online representation learning. Previously, I worked at TUKL-SEECS with Dr. Faisal Shafait on various Computer Vision and Machine Learning R&D projects during most of my undergraduate. I also represented my country, Pakistan, at 55th International Mathematical Olympiad , and XXVI Asian Pacific Mathematical Olympiad before college, receiving an honorable-mention and a bronze medal respectively.

CV / Google Scholar / Github / Twitter


Feb-2020: I moved to MILA to work with with Prof. Yoshua Bengio. I'll be working on meta-learning top-down modulation schemes for attention and plasticity. I'm also exploring the role of causal models in systematic generalization.

My Research

Learning Causal Models Online
K. Javed, M. White, Y.Bengio

We propose a method for learning models that do not rely on spurious correlations. Our work builds on IRM (M Arjovsky, 2019) except unlike IRM, it can be implemented online to (1) detect spurious features for a set of given features and (2) learn non-spurious features from sensory data.

Paper / Code

Meta-Learning Representations for Continual Learning
K. Javed and M. White

We propose OML, an objective for learning representations by using catastrophic interference as a training signal. Resultant representations are naturally sparse, accelerate future learning and are robust to forgetting under online updates in continual learning.

Paper / Code / Talk / Poster

Display Picture
Simultaneous Prediction Intervals for Patient-Specific Survival Curves
S. Sokota, R. D'Orazio, K. Javed, H. Haider and R. Greiner.

We propose a simple drop-in procedure for approximating the Bayesian credible regions of patient-specific survival functions that can be applied to many ISD models.

Paper / Code

Revisiting Distillation and Incremental Classifier Learning
K. Javed and F. Shafait

We isolate the truly effective existing ideas for incremental classifier learning from those that only work under certain conditions. Moreover, we propose a dynamic threshold moving algorithm that can successfully remove bias from an incrementally learned classifier when learning by knowledge distillation.

Paper / Poster / Code

Real-Time Document Localization in Natural Images by Recursive Application of a CNN (Oral)
K. Javed and F. Shafait

We propose a computationally efficient document segmentation algorithm that recursively uses convolutional neural networks to precisely localize a document in a natural image in real-time.

Paper / Slides / Code

Recent Talk

Khurram Javed. Design inspired by this website.