Karim Ahmed

K a r i m . A h m e d . P h D @ g m a i l . c o m
LinkedIn | Google Scholar | GitHub

About me

My research interests are in Deep Learning.
Ph.D. Thesis: Neural Connectivity Learning for Visual Recognition.


Research

Task Meta-Transfer from Limited Parallel Labels
Yiren Jian, Karim Ahmed, Lorenzo Torresani
NeurIPS (Meta-Learning Workshop) 2020
Meta-Learning Semantic Segmentation Depth Estimation Surface Normal Estimation 
Paper Poster
Star-Caps: Capsule Network with Straight-Through Attentive Routing
Karim Ahmed, Lorenzo Torresani
NeurIPS 2019
Capsule Networks Object Viewpoint Representation Self-Attention Connectivity Learning 
Paper Slides Poster
MaskConnect: Connectivity Learning by Gradient Descent
Karim Ahmed, Lorenzo Torresani
ECCV 2018
Neural Connectivity Learning AutoML 
Paper Summary Poster
Weighted Transformer Network for Machine Translation
Karim Ahmed, Nitish Shirish Keskar, Richard Socher
ArXiv 2017
Transformer Network Self-Attention Machine Translation 
Paper Blog
Connectivity Learning in Multi-Branch Networks
Karim Ahmed, Lorenzo Torresani
NIPS (Meta-learning Workshop) 2017
Nerual Connectivity Learning 
Paper Poster
BranchConnect: Large-Scale Visual Recognition with Learned Branch Connection
Karim Ahmed, Lorenzo Torresani
WACV 2018
Connectivity Learning Multi-Branch Neural Networks Large-Scale Image Classification 
Paper Slides Poster
Network of Experts for Large-Scale Image Categorization
Karim Ahmed, Mohamed Haris Baig, Lorenzo Torresani
ECCV 2016
Network of Experts Multi-Branch Neural Networks Large-Scale Image Classification 
Paper Poster Project

+ Other Publications


Academic Services

Reviewer: CVPR, ICCV, ECCV, TPAMI, NeurIPS.