free html web templates

Probability-based detection quality (pdq): A probabilistic approach to detection evaluation

New visual object detector evaluation measure which not only assesses detection quality, but also accounts for the spatial and label uncertainties produced by object detection systems.

A Theoretically Sound Upper Bound on the Triplet Loss for Improving the Efficiency of Deep Distance Metric Learning

In this paper, we first propose a solid theory on the linearization of the triplet loss with the use of class centroids, where the main conclusion is that our new linear loss represents a tight upper-bound to the triplet loss.

Bayesian Generative Active Deep Learning

In this paper, we propose a Bayesian generative active deep learning approach that combines active learning with data augmentation – we provide theoretical and empirical evidence (MNIST, CIFAR-{10, 100}, and SVHN) that our approach has more efficient training and better classification results than data augmentation and active learning.

Approximate Fisher Information Matrix to Characterise the Training of Deep Neural Networks.

In this paper, we introduce a novel methodology for characterising the performance of deep learning networks (ResNets and DenseNet) with respect to training convergence and generalisation as a function of mini-batch size and learning rate for image classification.

Bayesian Instance Segmentation in Open Set World

This paper addresses the semantic instance segmentation

task in the open-set conditions, where input images can contain known

and unknown object classes

Multi-modal Cycle-consistent Generalized Zero-Shot Learning

 In this paper, we propose the use of such constraint based on a new regularization for the GAN training that forces the generated visual features to reconstruct their original semantic features. Once our model is trained with this multi-modal cycle-consistent semantic compatibility, we can then synthesize more representative visual representations for the seen and, more importantly, for the unseen classes

A Bayesian Data Augmentation Approach for Learning Deep Models

In this work, we provide a novel Bayesian formulation to data augmentation, where new annotated training points are treated as missing variables and generated based on the distribution learned from the training set. 

A Deep Convolutional Neural Network Module that Promotes Competition of Multiple-size Filters

In this paper, we introduce a new module for deep ConvNets composed of several convolutional filters of multiple sizes that are joined by a maxout activation unit, which promotes competition amongst these filters. 

Learning Local Image Descriptors with Deep Siamese and Triplet Convolutional Networks by Minimising Global Loss Functions

We propose the use of a global loss that minimises the overall classification error in the training set, which can improve the generalisation capability of the model for learning local image descriptors. 

Unsupervised CNN for Single View Depth Estimation: Geometry to the Rescue

In this work we propose a unsupervised framework to learn a deep convolutional neural network for single view depth prediction, without requiring a pre-training stage or annotated ground-truth depths

Robust Optimization for Deep Regression

New robust loss function for the problem of 2D human pose estimation that allows for a significantly faster training.

On the Importance of Normalisation Layers in Deep Learning with Piecewise Linear Activation Units

We propose the introduction of batch normalisation units into deep feedforward neural networks with piecewise linear activations, which drives a more balanced use of these activation units