free website templates

Adapting Deep Learning for Real-world Medical Image Datasets

The project aims to investigate new deep learning modelling approaches to leverage real-world large-scale image data sets that contain noisy and incomplete labels and imbalanced class prevalence – to enable the use of these data sets for modelling deep learning classifiers. Expected outcomes include an innovative method for modelling deep learning classifiers. The research will involve new inter-disciplinary and international collaborations with machine learning and medical image analysis research institutions. This should provide significant benefits, such as better understanding of deep learning theory, new deep learning applications that can use previously unexplored data sets, and training for the future Australian workforce.

Selected Publications

Publications

Collaborators and Students

Workshops and Tutorials

Invited Seminars

Contact

PAC-Bayes Meta-learning with Implicit Task-specific Posteriors

We introduce a new and rigorously-formulated PAC-Bayes meta-learning algorithm that solves few-shot learning [PAMI'22 Paper]

Deep One-Class Classification via Interpolated Gaussian Descriptor

We introduce the interpolated Gaussian descriptor (IGD) method, a novel OCC model that

learns a one-class Gaussian anomaly classifier trained with adversarially interpolated training samples [AAAI'22 paper].[Presentation video]

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

PropMix filters out hard noisy samples, with the goal of increasing the likelihood of correctly re-labelling the easy noisy samples. Also, PropMix places clean and re-labelled easy noisy samples in a training set that is augmented with MixUp, removing the clean set size constraint and including a large proportion of correctly re-labelled easy noisy samples [BMVC'21 paper]

Perturbed and Strict Mean Teachers for Semi-supervised Semantic Segmentation

New semi-supervised segmentation with novel extensions of the mean-teacher (MT) model, which include a new auxiliary teacher, and the replacement of MT’s mean square error (MSE) by a stricter confidenceweighted cross-entropy (Conf-CE) loss [CVPR'22 paper].

ACPL: Anti-curriculum Pseudo-labelling for
Semi-supervised Medical Image Classification

New semi-supervised learning algorithm to select informative unlabelled samples, improving training balance and allowing the model to work for both multi-label and multi-class problems, and to estimate pseudo labels by an accurate ensemble of classifiers (improving pseudo label accuracy)                   [CVPR'22 Paper][Presentation Video].

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

New self-supervised representation learning method, called Constrained Contrastive Distribution learning for anomaly detection (CCD): fine-grained feature representations by simultaneously predicting the distribution of augmented data and image contexts using contrastive learning with pretext constraints [ICCV'21 Paper]

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

EvidentialMix: learning algorithm robust to a combination of closed-set and open-set noise [WACV'21 Paper]

Publications


  • Filipe R. Cordeiro,  Ragav Sachdeva, Vasileios Belagiannis, Ian Reid, and Gustavo Carneiro. "Longremix: Robust learning with high confidence samples in a noisy label environment." Pattern Recognition (2022): 109013. [pdf]
  • Cuong Nguyen, Thanh-Toan Do, and Gustavo Carneiro. "PAC-Bayes meta-learning with implicit task-specific posteriors." IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2022. [pdf]
  • Fengbei Liu, Yu Tian, Yuanhong Chen, Yuyuan Liu, Vasileios Belagiannis, and Gustavo Carneiro. "ACPL: Anti-curriculum Pseudo-labelling for Semi-supervised Medical Image Classification." Conference on Computer Vision and Pattern Recognition (CVPR) 2022. [pdf]
  • Yuyuan Liu, Yu Tian, Yuanhong Chen, Fengbei Liu, Vasileios Belagiannis, and Gustavo Carneiro. "Perturbed and Strict Mean Teachers for Semi-supervised Semantic Segmentation." Conference on Computer Vision and Pattern Recognition (CVPR) 2022. [pdf]
  • Yu Tian, Yuyuan Liu, Guansong Pang, Fengbei Liu, Yuanhong Chen, and Gustavo Carneiro. Pixel-wise Energy-biased Abstention Learning for Anomaly Segmentation on Complex Urban Driving Scenes. European Conference on Computer Vision (ECCV) 2022 [pdf]
  • Hu Wang, Jianpeng Zhang, Yuanhong Chen, Congbo Ma, Jodie Avery, Louise Hull, Gustavo Carneiro. Uncertainty-aware Multi-modal Learning via Cross-modal Random Network Prediction. European Conference on Computer Vision (ECCV) 2022 [pdf]
  • Fengbei Liu, Yuanhong Chen, Yu Tian, Yuyuan Liu, Chong Wang, Vasileios Belagiannis, Gustavo Carneiro. NVUM: Non-Volatile Unbiased Memory for Robust Medical Image Classification. International Conference Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2022. [pdf]
  • Yuanhong Chen, Hu Wang, Chong Wang, Yu Tian, Fengbei Liu, Yuyuan Liu, Michael Elliott, Davis McCarthy, Helen Frazer, Gustavo Carneiro. Multi-view Local Co-occurrence and Global Consistency Learning Improve Mammogram Classification Generalisation. International Conference Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2022. [pdf]
  • Yu Tian, Guansong Pang, Fengbei Liu, Yuyuan Liu, Chong Wang, Yuanhong Chen, Johan Verjans, Gustavo Carneiro. Contrastive Transformer-based Multiple Instance Learning for Weakly Supervised Polyp Frame Detection. International Conference Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2022. [pdf]
  • Chong Wang, Yuanhong Chen, Yuyuan Liu, Yu Tian, Fengbei Liu, Davis McCarthy, Michael Elliott, Helen Frazer, Gustavo Carneiro. Knowledge Distillation to Ensemble Global and Interpretable Prototype-based Mammogram Classification Models. International Conference Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2022. [pdf]
  • Renato Hermoza Aragones, Gabriel Maicas, Jacinto Nascimento, Gustavo Carneiro. Censor-aware Semi-supervised Learning for Survival Time Prediction from Medical Images. International Conference Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2022. [pdf]
  • Adrian Galdran, Narmin Ghaffari Laleh, Katherine Jane Hewitt, Jakob Kather, Miguel Angel González Ballester, Gustavo Carneiro. Test Time Transform Prediction for Open Set Histopathological Image Recognition. International Conference Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2022. [pdf]
  • Yuanhong Chen, Yu Tian, Guansong Pang, and Gustavo Carneiro. Unsupervised Anomaly Detection with Multi-scale Interpolated Gaussian Descriptors. AAAI Conference on Artificial Intelligence (AAAI-22).[pdf]
  • Emeson Santana, Gustavo Carneiro, and Filipe R. Cordeiro. "A Study on the Impact of Data Augmentation for Training Convolutional Neural Networks in the Presence of Noisy Labels." SIBGRAPI - Conference on Graphics, Patterns and Images (2022).[pdf]
  • David Butler, Yuan Zhang, Tim Chen, Seon Ho Shin, Rajvinder Singh, Gustavo Carneiro. In Defense of Kalman Filtering for Polyp Treacking from Colonoscopy Videos. International Symposium on Biomedical Imaging (ISBI) 2022.[pdf]
  • Carlos Santiago, Catarina Barata, Michele Sasdelli, Gustavo Carneiro, Jacinto Nascimento. LOW: Training Deep Neural Networks by Learning Optimal Sample Weights. Pattern Recognition (2021) [link]
  • Yu Tian, Guansong Pang, Yuanhong Chen, Rajvinder Singh, Johan W. Verjans, Gustavo Carneiro. Weakly-supervised Video Anomaly Detection with Robust Temporal Feature Magnitude Learning. International Conference on Computer Vision (ICCV) 2021. [pdf]
  • Filipe R. Cordeiro, Vasileios Belagiannis, Ian Reid, Gustavo Carneiro. PropMix: Hard Sample Filtering and Proportional MixUp for Learning with Noisy Labels. British Machine Vision Conference (BMVC) 2021.[pdf]
  • Cuong Nguyen, Thanh-Toan Do, Gustavo Carneiro. Probabilistic task modelling for meta-learning. Conference on Uncertainty in Artificial Intelligence (UAI) 2021 [pdf].
  • Yaqub Jonmohamadi, Shahnevaz Ali, Fengbei Liu, Jonathan Roberts, Ross Crawford, Gustavo Carneiro, Ajay Pandey. 3D Semantic Mapping from Arthroscopy using Out-of-distribution Pose and Depth and In-distribution Segmentation Training. International Conference Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2021. [pdf]
  • Minh-Son To, Ian Sarno, Chee Chong, Mark Jenkinson, Gustavo Carneiro. Self-supervised Lesion Change Detection and Localisation in Longitudinal Multiple Sclerosis Brain Imaging. International Conference Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2021. [pdf]
  • Yu Tian, Guansong Pang, Fengbei Liu, Seon Ho Shin, Johan W. Verjans, Rajvinder Singh, and Gustavo Carneiro. Constrained Contrastive Distribution Learning for Unsupervised Anomaly Detection and Localisation in Medical Images. International Conference Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2021. [pdf]
  • Adrian Galdran , Gustavo Carneiro, Miguel Angel González Ballester. Balanced-MixUp for highly imbalanced medical image classification. International Conference Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2021. [pdf]
  • Ragav Sachdeva, Filipe R. Cordeiro, Vasileios Belagiannis, Ian Reid, Gustavo Carneiro. EvidentialMix: Learning with Combined Open-set and Closed-set Noisy Labels. Winter Conference on Applications of Computer Vision (WACV) 2021.[pre-print]
  • Hoang Son Le, Rini Akmeliawati, Gustavo Carneiro (2021). Domain Generalisation with Domain Augmented Supervised Contrastive Learning. AAAI Conference on Artificial Intelligence. Student Abstract. 2021 [pdf]. 
  • Michele Sasdelli, Thalaiyasingam Ajanthan, Tat-Jun Chin, and Gustavo Carneiro. A Chaos Theory Approach to Understand Neural Network Optimization. International Conference on Digital Image Computing: Techniques and Applications (DICTA) 2021. [pdf]
  • Hoang Son Le, Rini Akmeliawati, Gustavo Carneiro. Combining Data Augmentation and Domain Distance Minimisation to Reduce Domain Generalisation Error. International Conference on Digital Image Computing: Techniques and Applications (DICTA) 2021. [pdf]
  • Adrian Johnston, Gustavo Carneiro. Self-supervised Monocular Trained Depth Estimation using Self-attention and Discrete Disparity Volume. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2020) [pre-print]
  • Yu Tian, Gabriel Maicas, Leonardo Zorron Cheng Tao Pu, Rajvinder Singh, Johan W. Verjans, Gustavo Carneiro. Few-Shot Anomaly Detection for Polyp Frames from Colonoscopy. In International Conference on Medical Imaging Computing and Computer-Assisted Intervention (MICCAI) 2020 [pre-print]
  • Youssef Dawoud, Julia Hornauer, Gustavo Carneiro, Vasileios Belagiannis. Few-Shot Microscopy Image Cell Segmentation. European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD) 2020 [pre-print]
  • Cuong Nguyen, Thanh-Toan Do, Gustavo Carneiro. Uncertainty in Model-Agnostic Meta-Learning using Variational Inference. Winter Conference on Applications of Computer Vision (WACV ’20) [pre-print]
  • Fabio Faria and Gustavo Carneiro. Why are Generative Adversarial Networks so Fascinating and Annoying?. 2020 33rd SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI). IEEE, 2020.[pre-print]
  • Filipe Cordeiro and Gustavo Carneiro. A Survey on Deep Learning with Noisy Labels: How to train your model when you cannot trust on the annotations? 2020 33rd SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI). IEEE, 2020. [pre-print]

Collaborators and Students

Ian Reid (Collaborator)

Vasileios Belagiannis (Collaborator)

Filipe Cordeiro (Collaborator)

Fabio Faria (Collaborator)

Jacinto Nascimento (Collaborator)

Cuong Cao Nguyen (Research Assistant)

Yu Tian (PhD student)

Fengbei Liu (PhD student)

Yuyuan Liu (PhD student)

Yuanhong Chen (PhD student)

Arpit Garg (PhD student)

Dung Anh Hoang (MPhil student)

Brandon Smart (Graduated in Dec'2020 - Honours. Now a research intern until July 2022)

Ragav Sachdeva (Graduated in Dec'2020 - Honours)

Workshops and Tutorials


Invited Seminars


ADDRESS
Australian Institute for Machine Learning, Lot Fourteen
University of Adelaide
Adelaide 5005, Australia

CONTACTS
gustavo.carneiro@adelaide.edu.au