Abstract Representation learning plays a key role in building robust and general-purpose
vision learners, and is a long-standing problem. It becomes increasingly interesting with the …

This paper deals with unsupervised collection of information from traffic surveillance video
streams. Deployment of usable traffic surveillance systems requires minimizing of efforts per …

We propose learning to transfer learn (L2TL) to improve transfer learning on a target dataset
by judicious extraction of information from a source dataset. L2TL considers joint …

The massive growth in the number of vehicles has been noticed over the past few decades
with the increase in population all over the world. Therefore, tracking of vehicles depending …

This thesis considers deep learning theories of brain function, and in particular biologically
plausible deep learning. The idea is to treat a standard deep network as a high-level model …

Prior works on self-supervised pre-training focus on the joint training scenario, where
massive unlabeled data are assumed to be given as input all at once, and only then is a …

In this thesis we demonstrate the benefits of automated labeled dataset creation for fine-
grained visual learning tasks. Specifically, we show that utilizing real-world, non-image …

The purpose of this dissertation in the field of Computer Science is to showcase the
development process of a deep learning algorithm, used to classify the make and model of a …

3. Extended experimental results 1 3.1. Impact of data augmentation......... 1 3.2. Results on
the ImageNet-CoG [ 25] benchmark 2 3.3. Analysis of the learned features....... 3 3.4. Impact …

Vehicle detection and recognition is the research focus in Intelligent Transportation System
(ITS) with many challenges. Based on the success of Convolutional neural networks (CNN) …