Search Results for author: Gustavo Carneiro

Found 65 papers, 25 papers with code

PropMix: Hard Sample Filtering and Proportional MixUp for Learning with Noisy Labels

1 code implementation22 Oct 2021 Filipe R. Cordeiro, Vasileios Belagiannis, Ian Reid, Gustavo Carneiro

The most competitive noisy label learning methods rely on an unsupervised classification of clean and noisy samples, where samples classified as noisy are re-labelled and "MixMatched" with the clean samples.

Learning with noisy labels

Balanced-MixUp for Highly Imbalanced Medical Image Classification

1 code implementation20 Sep 2021 Adrian Galdran, Gustavo Carneiro, Miguel A. González Ballester

The resulting two sets of samples are then mixed-up to create a more balanced training distribution from which a neural network can effectively learn without incurring in heavily under-fitting the minority classes.

Classification Image Classification

Multi-centred Strong Augmentation via Contrastive Learning for Unsupervised Lesion Detection and Segmentation

1 code implementation3 Sep 2021 Yu Tian, Fengbei Liu, Guansong Pang, Yuanhong Chen, Yuyuan Liu, Johan W. Verjans, Rajvinder Singh, Gustavo Carneiro

MSACL learns representations by separating several types of strong and weak augmentations of normal image samples, where the weak augmentations represent normal images and strong augmentations denote synthetic abnormal images.

Contrastive Learning Data Augmentation +1

Probabilistic task modelling for meta-learning

1 code implementation9 Jun 2021 Cuong C. Nguyen, Thanh-Toan Do, Gustavo Carneiro

We propose probabilistic task modelling -- a generative probabilistic model for collections of tasks used in meta-learning.

Meta-Learning Variational Inference

Self-supervised Lesion Change Detection and Localisation in Longitudinal Multiple Sclerosis Brain Imaging

no code implementations2 Jun 2021 Minh-Son To, Ian G Sarno, Chee Chong, Mark Jenkinson, Gustavo Carneiro

Hence, we introduce a new unsupervised anomaly detection and localisation method trained exclusively with serial images that do not contain any lesion changes.

Decision Making Unsupervised Anomaly Detection

ScanMix: Learning from Severe Label Noise via Semantic Clustering and Semi-Supervised Learning

1 code implementation21 Mar 2021 Ragav Sachdeva, Filipe R Cordeiro, Vasileios Belagiannis, Ian Reid, Gustavo Carneiro

To be specific, ScanMix is designed based on the expectation maximisation (EM) framework, where the E-step estimates the value of a latent variable to cluster the training images based on their appearance representations and classification results, and the M-step optimises the SSL classification and learns effective feature representations via semantic clustering.

Image Classification

LongReMix: Robust Learning with High Confidence Samples in a Noisy Label Environment

no code implementations6 Mar 2021 Filipe R. Cordeiro, Ragav Sachdeva, Vasileios Belagiannis, Ian Reid, Gustavo Carneiro

Deep neural network models are robust to a limited amount of label noise, but their ability to memorise noisy labels in high noise rate problems is still an open issue.

Image Classification

Noisy Label Learning for Large-scale Medical Image Classification

1 code implementation6 Mar 2021 Fengbei Liu, Yu Tian, Filipe R. Cordeiro, Vasileios Belagiannis, Ian Reid, Gustavo Carneiro

The classification accuracy of deep learning models depends not only on the size of their training sets, but also on the quality of their labels.

Classification General Classification +1

Constrained Contrastive Distribution Learning for Unsupervised Anomaly Detection and Localisation in Medical Images

1 code implementation5 Mar 2021 Yu Tian, Guansong Pang, Fengbei Liu, Yuanhong Chen, Seon Ho Shin, Johan W. Verjans, Rajvinder Singh, Gustavo Carneiro

Unsupervised anomaly detection (UAD) learns one-class classifiers exclusively with normal (i. e., healthy) images to detect any abnormal (i. e., unhealthy) samples that do not conform to the expected normal patterns.

Contrastive Learning Representation Learning +1

Self-supervised Mean Teacher for Semi-supervised Chest X-ray Classification

1 code implementation5 Mar 2021 Fengbei Liu, Yu Tian, Filipe R. Cordeiro, Vasileios Belagiannis, Ian Reid, Gustavo Carneiro

In this paper, we propose Self-supervised Mean Teacher for Semi-supervised (S$^2$MTS$^2$) learning that combines self-supervised mean-teacher pre-training with semi-supervised fine-tuning.

Classification Contrastive Learning +3

Post-hoc Overall Survival Time Prediction from Brain MRI

1 code implementation22 Feb 2021 Renato Hermoza, Gabriel Maicas, Jacinto C. Nascimento, Gustavo Carneiro

In this paper, we introduce a new post-hoc method for OS time prediction that does not require segmentation map annotation for training.

Brain Tumor Segmentation Tumor Segmentation

Similarity of Classification Tasks

1 code implementation27 Jan 2021 Cuong Nguyen, Thanh-Toan Do, Gustavo Carneiro

Recent advances in meta-learning has led to remarkable performances on several few-shot learning benchmarks.

Classification Few-Shot Learning +1

Deep One-Class Classification via Interpolated Gaussian Descriptor

no code implementations25 Jan 2021 Yuanhong Chen, Yu Tian, Guansong Pang, Gustavo Carneiro

The adversarial interpolation is enforced to consistently learn a smooth Gaussian descriptor, even when the training data is small or contaminated with anomalous samples.

Classification Unsupervised Anomaly Detection

Weakly-supervised Video Anomaly Detection with Robust Temporal Feature Magnitude Learning

2 code implementations ICCV 2021 Yu Tian, Guansong Pang, Yuanhong Chen, Rajvinder Singh, Johan W. Verjans, Gustavo Carneiro

To address this issue, we introduce a novel and theoretically sound method, named Robust Temporal Feature Magnitude learning (RTFM), which trains a feature magnitude learning function to effectively recognise the positive instances, substantially improving the robustness of the MIL approach to the negative instances from abnormal videos.

Anomaly Detection In Surveillance Videos Contrastive Learning +1

Semantics for Robotic Mapping, Perception and Interaction: A Survey

no code implementations2 Jan 2021 Sourav Garg, Niko Sünderhauf, Feras Dayoub, Douglas Morrison, Akansel Cosgun, Gustavo Carneiro, Qi Wu, Tat-Jun Chin, Ian Reid, Stephen Gould, Peter Corke, Michael Milford

In robotics and related research fields, the study of understanding is often referred to as semantics, which dictates what does the world "mean" to a robot, and is strongly tied to the question of how to represent that meaning.

Autonomous Driving Human robot interaction

A Chaos Theory Approach to Understand Neural Network Optimization

no code implementations1 Jan 2021 Michele Sasdelli, Thalaiyasingam Ajanthan, Tat-Jun Chin, Gustavo Carneiro

Then, we empirically show that for a large range of learning rates, SGD traverses the loss landscape across regions with largest eigenvalue of the Hessian similar to the inverse of the learning rate.

Bayesian Metric Learning for Robust Training of Deep Models under Noisy Labels

no code implementations1 Jan 2021 Toan Tran, Hieu Vu, Gustavo Carneiro, Hung Bui

Label noise is a natural event of data collection and annotation and has been shown to have significant impact on the performance of deep learning models regarding accuracy reduction and sample complexity increase.

Classification General Classification +2

Domain Generalisation with Domain Augmented Supervised Contrastive Learning (Student Abstract)

no code implementations27 Dec 2020 Hoang Son Le, Rini Akmeliawati, Gustavo Carneiro

Domain generalisation (DG) methods address the problem of domain shift, when there is a mismatch between the distributions of training and target domains.

Contrastive Learning Data Augmentation

A Survey on Deep Learning with Noisy Labels: How to train your model when you cannot trust on the annotations?

no code implementations5 Dec 2020 Filipe R. Cordeiro, Gustavo Carneiro

As deep learning models depend on correctly labeled data sets and label correctness is difficult to guarantee, it is crucial to consider the presence of noisy labels for deep learning training.

Learning with noisy labels Meta-Learning

EvidentialMix: Learning with Combined Open-set and Closed-set Noisy Labels

1 code implementation11 Nov 2020 Ragav Sachdeva, Filipe R. Cordeiro, Vasileios Belagiannis, Ian Reid, Gustavo Carneiro

In this work, we study a new variant of the noisy label problem that combines the open-set and closed-set noisy labels, and introduce a benchmark evaluation to assess the performance of training algorithms under this setup.

Deep Metric Learning Meets Deep Clustering: An Novel Unsupervised Approach for Feature Embedding

1 code implementation9 Sep 2020 Binh X. Nguyen, Binh D. Nguyen, Gustavo Carneiro, Erman Tjiputra, Quang D. Tran, Thanh-Toan Do

Based on pseudo labels, we propose a novel unsupervised metric loss which enforces the positive concentration and negative separation of samples in the embedding space.

Deep Clustering Metric Learning

Self-supervised Depth Estimation to Regularise Semantic Segmentation in Knee Arthroscopy

no code implementations5 Jul 2020 Fengbei Liu, Yaqub Jonmohamadi, Gabriel Maicas, Ajay K. Pandey, Gustavo Carneiro

In this paper, we propose a novel self-supervised monocular depth estimation to regularise the training of the semantic segmentation in knee arthroscopy.

Monocular Depth Estimation Semantic Segmentation

Few-Shot Microscopy Image Cell Segmentation

1 code implementation29 Jun 2020 Youssef Dawoud, Julia Hornauer, Gustavo Carneiro, Vasileios Belagiannis

Instead, we assume that we can access a plethora of annotated image data sets from different domains (sources) and a limited number of annotated image data sets from the domain of interest (target), where each domain denotes not only different image appearance but also a different type of cell segmentation problem.

Cell Segmentation Few-Shot Learning

Few-Shot Anomaly Detection for Polyp Frames from Colonoscopy

1 code implementation26 Jun 2020 Yu Tian, Gabriel Maicas, Leonardo Zorron Cheng Tao Pu, Rajvinder Singh, Johan W. Verjans, Gustavo Carneiro

Anomaly detection methods generally target the learning of a normal image distribution (i. e., inliers showing healthy cases) and during testing, samples relatively far from the learned distribution are classified as anomalies (i. e., outliers showing disease cases).

Few Shot Anomaly Detection

Region Proposals for Saliency Map Refinement for Weakly-supervised Disease Localisation and Classification

1 code implementation21 May 2020 Renato Hermoza, Gabriel Maicas, Jacinto C. Nascimento, Gustavo Carneiro

The deployment of automated systems to diagnose diseases from medical images is challenged by the requirement to localise the diagnosed diseases to justify or explain the classification decision.

Classification General Classification +2

Self-supervised Monocular Trained Depth Estimation using Self-attention and Discrete Disparity Volume

1 code implementation CVPR 2020 Adrian Johnston, Gustavo Carneiro

We show that the extension of the state-of-the-art self-supervised monocular trained depth estimator Monodepth2 with these two ideas allows us to design a model that produces the best results in the field in KITTI 2015 and Make3D, closing the gap with respect self-supervised stereo training and fully supervised approaches.

Monocular Depth Estimation

PAC-Bayesian Meta-learning with Implicit Prior and Posterior

no code implementations5 Mar 2020 Cuong Nguyen, Thanh-Toan Do, Gustavo Carneiro

We introduce a new and rigorously-formulated PAC-Bayes few-shot meta-learning algorithm that implicitly learns a prior distribution of the model of interest.

Classification Few-Shot Image Classification +1

Unsupervised Dual Adversarial Learning for Anomaly Detection in Colonoscopy Video Frames

no code implementations23 Oct 2019 Yuyuan Liu, Yu Tian, Gabriel Maicas, Leonardo Z. C. T. Pu, Rajvinder Singh, Johan W. Verjans, Gustavo Carneiro

We show that our proposed approach achieves the state-of-the-art result on this data set, compared with recently proposed anomaly detection systems.

Anomaly Detection

Semi-supervised Multi-domain Multi-task Training for Metastatic Colon Lymph Node Diagnosis From Abdominal CT

no code implementations23 Oct 2019 Saskia Glaser, Gabriel Maicas, Sergei Bedrikovetski, Tarik Sammour, Gustavo Carneiro

However, the lack of annotations for the localisation of the regions of interest (ROIs) containing lymph nodes can limit classification accuracy due to the small size of the relevant ROIs in this problem.

Computed Tomography (CT) General Classification

Hidden Stratification Causes Clinically Meaningful Failures in Machine Learning for Medical Imaging

no code implementations27 Sep 2019 Luke Oakden-Rayner, Jared Dunnmon, Gustavo Carneiro, Christopher Ré

Machine learning models for medical image analysis often suffer from poor performance on important subsets of a population that are not identified during training or testing.

Generalised Zero-Shot Learning with Domain Classification in a Joint Semantic and Visual Space

no code implementations14 Aug 2019 Rafael Felix, Ben Harwood, Michele Sasdelli, Gustavo Carneiro

Most state-of-the-art GZSL approaches rely on a mapping between latent visual and semantic spaces without considering if a particular sample belongs to the set of seen or unseen classes.

Classification General Classification +1

Few-Shot Meta-Denoising

no code implementations31 Jul 2019 Leslie Casas, Attila Klimmek, Gustavo Carneiro, Nassir Navab, Vasileios Belagiannis

A solution to mitigate the small training set issue is to pre-train a denoising model with small training sets containing pairs of clean and synthesized noisy signals, produced from empirical noise priors, and fine-tune on the available small training set.

Denoising Few-Shot Learning +1

Uncertainty in Model-Agnostic Meta-Learning using Variational Inference

1 code implementation27 Jul 2019 Cuong Nguyen, Thanh-Toan Do, Gustavo Carneiro

We introduce a new, rigorously-formulated Bayesian meta-learning algorithm that learns a probability distribution of model parameter prior for few-shot learning.

Classification Few-Shot Image Classification +2

Unsupervised Task Design to Meta-Train Medical Image Classifiers

no code implementations17 Jul 2019 Gabriel Maicas, Cuong Nguyen, Farbod Motlagh, Jacinto C. Nascimento, Gustavo Carneiro

Meta-training has been empirically demonstrated to be the most effective pre-training method for few-shot learning of medical image classifiers (i. e., classifiers modeled with small training sets).

Classification Few-Shot Learning +1

Bayesian Generative Active Deep Learning

no code implementations26 Apr 2019 Toan Tran, Thanh-Toan Do, Ian Reid, Gustavo Carneiro

Deep learning models have demonstrated outstanding performance in several problems, but their training process tends to require immense amounts of computational and human resources for training and labeling, constraining the types of problems that can be tackled.

Active Learning Data Augmentation

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

no code implementations CVPR 2019 Thanh-Toan Do, Toan Tran, Ian Reid, Vijay Kumar, Tuan Hoang, Gustavo Carneiro

Another approach explored in the field relies on an ad-hoc linearization (in terms of N) of the triplet loss that introduces class centroids, which must be optimized using the whole training set for each mini-batch - this means that a naive implementation of this approach has run-time complexity O(N^2).

Metric Learning

Multi-modal Ensemble Classification for Generalized Zero Shot Learning

no code implementations15 Jan 2019 Rafael Felix, Michele Sasdelli, Ian Reid, Gustavo Carneiro

In this paper, we mitigate these issues by proposing a new GZSL method based on multi-modal training and testing processes, where the optimization explicitly promotes a balanced classification accuracy between seen and unseen classes.

Bayesian Inference Classification +2

Probabilistic Object Detection: Definition and Evaluation

1 code implementation27 Nov 2018 David Hall, Feras Dayoub, John Skinner, Haoyang Zhang, Dimity Miller, Peter Corke, Gustavo Carneiro, Anelia Angelova, Niko Sünderhauf

We introduce Probabilistic Object Detection, the task of detecting objects in images and accurately quantifying the spatial and semantic uncertainties of the detections.

Object Detection

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

1 code implementation16 Oct 2018 Zhibin Liao, Tom Drummond, Ian Reid, Gustavo Carneiro

Furthermore, the proposed measurements also allow us to show that it is possible to optimise the training process with a new dynamic sampling training approach that continuously and automatically change the mini-batch size and learning rate during the training process.

General Classification Image Classification

Pre and Post-hoc Diagnosis and Interpretation of Malignancy from Breast DCE-MRI

no code implementations25 Sep 2018 Gabriel Maicas, Andrew P. Bradley, Jacinto C. Nascimento, Ian Reid, Gustavo Carneiro

Conversely, traditional approaches follow a pre-hoc approach that initially localises suspicious areas that are subsequently classified to establish the breast malignancy -- this approach is trained using strongly annotated data (i. e., it needs a delineation and classification of all lesions in an image).

Multi-modal Cycle-consistent Generalized Zero-Shot Learning

1 code implementation ECCV 2018 Rafael Felix, B. G. Vijay Kumar, Ian Reid, Gustavo Carneiro

In generalized zero shot learning (GZSL), the set of classes are split into seen and unseen classes, where training relies on the semantic features of the seen and unseen classes and the visual representations of only the seen classes, while testing uses the visual representations of the seen and unseen classes.

General Classification Generalized Zero-Shot Learning

Model Agnostic Saliency for Weakly Supervised Lesion Detection from Breast DCE-MRI

no code implementations20 Jul 2018 Gabriel Maicas, Gerard Snaauw, Andrew P. Bradley, Ian Reid, Gustavo Carneiro

There is a heated debate on how to interpret the decisions provided by deep learning models (DLM), where the main approaches rely on the visualization of salient regions to interpret the DLM classification process.

Classification General Classification

Bayesian Semantic Instance Segmentation in Open Set World

no code implementations ECCV 2018 Trung Pham, Vijay Kumar B G, Thanh-Toan Do, Gustavo Carneiro, Ian Reid

In this paper, we present a novel open-set semantic instance segmentation approach capable of segmenting all known and unknown object classes in images, based on the output of an object detector trained on known object classes.

Instance Segmentation Semantic Segmentation

Producing radiologist-quality reports for interpretable artificial intelligence

no code implementations1 Jun 2018 William Gale, Luke Oakden-Rayner, Gustavo Carneiro, Andrew P. Bradley, Lyle J. Palmer

Current approaches to explaining the decisions of deep learning systems for medical tasks have focused on visualising the elements that have contributed to each decision.

Decision Making

Training Medical Image Analysis Systems like Radiologists

no code implementations28 May 2018 Gabriel Maicas, Andrew P. Bradley, Jacinto C. Nascimento, Ian Reid, Gustavo Carneiro

This process bears no direct resemblance with radiologist training, which is based on solving a series of tasks of increasing difficulty, where each task involves the use of significantly smaller datasets than those used in machine learning.

Classification Curriculum Learning +3

Detecting hip fractures with radiologist-level performance using deep neural networks

no code implementations17 Nov 2017 William Gale, Luke Oakden-Rayner, Gustavo Carneiro, Andrew P. Bradley, Lyle J. Palmer

We developed an automated deep learning system to detect hip fractures from frontal pelvic x-rays, an important and common radiological task.

Smart Mining for Deep Metric Learning

no code implementations ICCV 2017 Ben Harwood, Vijay Kumar B G, Gustavo Carneiro, Ian Reid, Tom Drummond

In this paper, we propose a novel deep metric learning method that combines the triplet model and the global structure of the embedding space.

Metric Learning

Automated Detection of Individual Micro-calcifications from Mammograms using a Multi-stage Cascade Approach

no code implementations7 Oct 2016 Zhi Lu, Gustavo Carneiro, Neeraj Dhungel, Andrew P. Bradley

In mammography, the efficacy of computer-aided detection methods depends, in part, on the robust localisation of micro-calcifications ($\mu$C).

General Classification

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

2 code implementations16 Mar 2016 Ravi Garg, Vijay Kumar BG, Gustavo Carneiro, Ian Reid

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

Depth Estimation

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

2 code implementations CVPR 2016 Vijay Kumar B G, Gustavo Carneiro, Ian Reid

Current results from machine learning show that replacing this siamese by a triplet network can improve the classification accuracy in several problems, but this has yet to be demonstrated for local image descriptor learning.

General Classification

Weakly-Supervised Structured Output Learning With Flexible and Latent Graphs Using High-Order Loss Functions

no code implementations ICCV 2015 Gustavo Carneiro, Tingying Peng, Christine Bayer, Nassir Navab

We introduce two new structured output models that use a latent graph, which is flexible in terms of the number of nodes and structure, where the training process minimises a high-order loss function using a weakly annotated training set.

Competitive Multi-scale Convolution

no code implementations18 Nov 2015 Zhibin Liao, Gustavo Carneiro

In this paper, we introduce a new deep convolutional neural network (ConvNet) module that promotes competition among a set of multi-scale convolutional filters.

Image Classification

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

no code implementations3 Aug 2015 Zhibin Liao, Gustavo Carneiro

The combination of deep learning models and piecewise linear activation functions allows for the estimation of exponentially complex functions with the use of a large number of subnetworks specialized in the classification of similar input examples.

General Classification Image Classification

Robust Optimization for Deep Regression

1 code implementation ICCV 2015 Vasileios Belagiannis, Christian Rupprecht, Gustavo Carneiro, Nassir Navab

Convolutional Neural Networks (ConvNets) have successfully contributed to improve the accuracy of regression-based methods for computer vision tasks such as human pose estimation, landmark localization, and object detection.

Age Estimation Object Detection +1

Deep Structured learning for mass segmentation from Mammograms

no code implementations27 Oct 2014 Neeraj Dhungel, Gustavo Carneiro, Andrew P. Bradley

In this paper, we present a novel method for the segmentation of breast masses from mammograms exploring structured and deep learning.

Mass Segmentation From Mammograms

Non-rigid Segmentation using Sparse Low Dimensional Manifolds and Deep Belief Networks

no code implementations CVPR 2014 Jacinto C. Nascimento, Gustavo Carneiro

In this paper, we propose a new methodology for segmenting non-rigid visual objects, where the search procedure is onducted directly on a sparse low-dimensional manifold, guided by the classification results computed from a deep belief network.

Top-Down Segmentation of Non-rigid Visual Objects Using Derivative-Based Search on Sparse Manifolds

no code implementations CVPR 2013 Jacinto C. Nascimento, Gustavo Carneiro

In this paper, we propose the use of sparse manifolds to reduce the dimensionality of the rigid detection search space of current stateof-the-art top-down segmentation methodologies.

Cannot find the paper you are looking for? You can Submit a new open access paper.