Change search
Link to record
Permanent link

Direct link
BETA
Publications (10 of 16) Show all publications
Englesson, E. & Azizpour, H. (2019). Efficient Evaluation-Time Uncertainty Estimation by Improved Distillation. In: : . Paper presented at International Conference on Machine Learning (ICML) Workshops, 2019 Workshop on Uncertainty and Robustness in Deep Learning.
Open this publication in new window or tab >>Efficient Evaluation-Time Uncertainty Estimation by Improved Distillation
2019 (English)Conference paper, Poster (with or without abstract) (Refereed)
National Category
Computer Vision and Robotics (Autonomous Systems)
Identifiers
urn:nbn:se:kth:diva-260511 (URN)
Conference
International Conference on Machine Learning (ICML) Workshops, 2019 Workshop on Uncertainty and Robustness in Deep Learning
Note

QC 20191001

Available from: 2019-09-30 Created: 2019-09-30 Last updated: 2019-10-01Bibliographically approved
Baldassarre, F. & Azizpour, H. (2019). Explainability Techniques for Graph Convolutional Networks. In: : . Paper presented at International Conference on Machine Learning (ICML) Workshops, 2019 Workshop on Learning and Reasoning with Graph-Structured Representations.
Open this publication in new window or tab >>Explainability Techniques for Graph Convolutional Networks
2019 (English)Conference paper, Poster (with or without abstract) (Refereed)
Abstract [en]

Graph Networks are used to make decisions in potentially complex scenarios but it is usually not obvious how or why they made them. In this work, we study the explainability of Graph Network decisions using two main classes of techniques, gradient-based and decomposition-based, on a toy dataset and a chemistry task. Our study sets the ground for future development as well as application to real-world problems.

National Category
Computer Vision and Robotics (Autonomous Systems)
Identifiers
urn:nbn:se:kth:diva-260507 (URN)
Conference
International Conference on Machine Learning (ICML) Workshops, 2019 Workshop on Learning and Reasoning with Graph-Structured Representations
Note

QC 20191001

Available from: 2019-09-30 Created: 2019-09-30 Last updated: 2019-10-01Bibliographically approved
Srinivasan, P. A., Guastoni, L., Azizpour, H., Schlatter, P. & Vinuesa, R. (2019). Predictions of turbulent shear flows using deep neural networks. Physical Review Fluids, 4(5), Article ID 054603.
Open this publication in new window or tab >>Predictions of turbulent shear flows using deep neural networks
Show others...
2019 (English)In: Physical Review Fluids, E-ISSN 2469-990X, Vol. 4, no 5, article id 054603Article in journal (Refereed) Published
Abstract [en]

In the present work, we assess the capabilities of neural networks to predict temporally evolving turbulent flows. In particular, we use the nine-equation shear flow model by Moehlis et al. [New J. Phys. 6, 56 (2004)] to generate training data for two types of neural networks: the multilayer perceptron (MLP) and the long short-term memory (LSTM) networks. We tested a number of neural network architectures by varying the number of layers, number of units per layer, dimension of the input, and weight initialization and activation functions in order to obtain the best configurations for flow prediction. Because of its ability to exploit the sequential nature of the data, the LSTM network outperformed the MLP. The LSTM led to excellent predictions of turbulence statistics (with relative errors of 0.45% and 2.49% in mean and fluctuating quantities, respectively) and of the dynamical behavior of the system (characterized by Poincare maps and Lyapunov exponents). This is an exploratory study where we consider a low-order representation of near-wall turbulence. Based on the present results, the proposed machine-learning framework may underpin future applications aimed at developing accurate and efficient data-driven subgrid-scale models for large-eddy simulations of more complex wall-bounded turbulent flows, including channels and developing boundary layers.

Place, publisher, year, edition, pages
AMER PHYSICAL SOC, 2019
National Category
Physical Sciences
Identifiers
urn:nbn:se:kth:diva-252606 (URN)10.1103/PhysRevFluids.4.054603 (DOI)000467744500004 ()
Note

QC 20190610

Available from: 2019-06-10 Created: 2019-06-10 Last updated: 2019-06-10Bibliographically approved
Robertson, S., Azizpour, H., Smith, K. & Hartman, J. (2018). Digital image analysis in breast pathology-from image processing techniques to artificial intelligence. Translational Research: The Journal of Laboratory and Clinical Medicine, 194, 19-35
Open this publication in new window or tab >>Digital image analysis in breast pathology-from image processing techniques to artificial intelligence
2018 (English)In: Translational Research: The Journal of Laboratory and Clinical Medicine, ISSN 1931-5244, E-ISSN 1878-1810, Vol. 194, p. 19-35Article, review/survey (Refereed) Published
Abstract [en]

Breast cancer is the most common malignant disease in women worldwide. In recent decades, earlier diagnosis and better adjuvant therapy have substantially improved patient outcome. Diagnosis by histopathology has proven to be instrumental to guide breast cancer treatment, but new challenges have emerged as our increasing understanding of cancer over the years has revealed its complex nature. As patient demand for personalized breast cancer therapy grows, we face an urgent need for more precise biomarker assessment and more accurate histopathologic breast cancer diagnosis to make better therapy decisions. The digitization of pathology data has opened the door to faster, more reproducible, and more precise diagnoses through computerized image analysis. Software to assist diagnostic breast pathology through image processing techniques have been around for years. But recent breakthroughs in artificial intelligence (AI) promise to fundamentally change the way we detect and treat breast cancer in the near future. Machine learning, a subfield of AI that applies statistical methods to learn from data, has seen an explosion of interest in recent years because of its ability to recognize patterns in data with less need for human instruction. One technique in particular, known as deep learning, has produced groundbreaking results in many important problems including image classification and speech recognition. In this review, we will cover the use of AI and deep learning in diagnostic breast pathology, and other recent developments in digital image analysis.

Place, publisher, year, edition, pages
ELSEVIER SCIENCE INC, 2018
National Category
Cancer and Oncology Radiology, Nuclear Medicine and Medical Imaging
Identifiers
urn:nbn:se:kth:diva-226196 (URN)10.1016/j.trsl.2017.10.010 (DOI)000428608600002 ()29175265 (PubMedID)2-s2.0-85036635168 (Scopus ID)
Note

QC 20180518

Available from: 2018-05-18 Created: 2018-05-18 Last updated: 2019-09-18Bibliographically approved
Smith, K., Piccinini, F., Balassa, T., Koos, K., Danka, T., Azizpour, H. & Horvath, P. (2018). Phenotypic Image Analysis Software Tools for Exploring and Understanding Big Image Data from Cell-Based Assays. CELL SYSTEMS, 6(6), 636-653
Open this publication in new window or tab >>Phenotypic Image Analysis Software Tools for Exploring and Understanding Big Image Data from Cell-Based Assays
Show others...
2018 (English)In: CELL SYSTEMS, ISSN 2405-4712, Vol. 6, no 6, p. 636-653Article, review/survey (Refereed) Published
Abstract [en]

Phenotypic image analysis is the task of recognizing variations in cell properties using microscopic image data. These variations, produced through a complex web of interactions between genes and the environment, may hold the key to uncover important biological phenomena or to understand the response to a drug candidate. Today, phenotypic analysis is rarely performed completely by hand. The abundance of high-dimensional image data produced by modern high-throughput microscopes necessitates computational solutions. Over the past decade, a number of software tools have been developed to address this need. They use statistical learning methods to infer relationships between a cell's phenotype and data from the image. In this review, we examine the strengths and weaknesses of non-commercial phenotypic image analysis software, cover recent developments in the field, identify challenges, and give a perspective on future possibilities.

Place, publisher, year, edition, pages
Elsevier, 2018
National Category
Bioinformatics and Systems Biology
Identifiers
urn:nbn:se:kth:diva-232249 (URN)10.1016/j.cels.2018.06.001 (DOI)000436877800002 ()29953863 (PubMedID)2-s2.0-85048445198 (Scopus ID)
Funder
Science for Life Laboratory - a national resource center for high-throughput molecular bioscience
Note

QC 20180720

Available from: 2018-07-20 Created: 2018-07-20 Last updated: 2019-09-17Bibliographically approved
Carlsson, S., Azizpour, H., Razavian, A. S., Sullivan, J. & Smith, K. (2017). The Preimage of Rectifier Network Activities. In: International Conference on Learning Representations (ICLR): . Paper presented at International Conference on Learning Representations (ICLR).
Open this publication in new window or tab >>The Preimage of Rectifier Network Activities
Show others...
2017 (English)In: International Conference on Learning Representations (ICLR), 2017Conference paper, Published paper (Refereed)
Abstract [en]

The preimage of the activity at a certain level of a deep network is the set of inputs that result in the same node activity. For fully connected multi layer rectifier networks we demonstrate how to compute the preimages of activities at arbitrary levels from knowledge of the parameters in a deep rectifying network. If the preimage set of a certain activity in the network contains elements from more than one class it means that these classes are irreversibly mixed. This implies that preimage sets which are piecewise linear manifolds are building blocks for describing the input manifolds specific classes, ie all preimages should ideally be from the same class. We believe that the knowledge of how to compute preimages will be valuable in understanding the efficiency displayed by deep learning networks and could potentially be used in designing more efficient training algorithms.

National Category
Computer Vision and Robotics (Autonomous Systems) Computer Vision and Robotics (Autonomous Systems)
Identifiers
urn:nbn:se:kth:diva-259164 (URN)2-s2.0-85071123889 (Scopus ID)
Conference
International Conference on Learning Representations (ICLR)
Note

QC 20190916

Available from: 2019-09-11 Created: 2019-09-11 Last updated: 2019-09-16Bibliographically approved
Azizpour, H., Sharif Razavian, A., Sullivan, J., Maki, A. & Carlssom, S. (2016). Factors of Transferability for a Generic ConvNet Representation. IEEE Transaction on Pattern Analysis and Machine Intelligence, 38(9), 1790-1802, Article ID 7328311.
Open this publication in new window or tab >>Factors of Transferability for a Generic ConvNet Representation
Show others...
2016 (English)In: IEEE Transaction on Pattern Analysis and Machine Intelligence, ISSN 0162-8828, E-ISSN 1939-3539, Vol. 38, no 9, p. 1790-1802, article id 7328311Article in journal (Refereed) Published
Abstract [en]

Evidence is mounting that Convolutional Networks (ConvNets) are the most effective representation learning method for visual recognition tasks. In the common scenario, a ConvNet is trained on a large labeled dataset (source) and the feed-forward units activation of the trained network, at a certain layer of the network, is used as a generic representation of an input image for a task with relatively smaller training set (target). Recent studies have shown this form of representation transfer to be suitable for a wide range of target visual recognition tasks. This paper introduces and investigates several factors affecting the transferability of such representations. It includes parameters for training of the source ConvNet such as its architecture, distribution of the training data, etc. and also the parameters of feature extraction such as layer of the trained ConvNet, dimensionality reduction, etc. Then, by optimizing these factors, we show that significant improvements can be achieved on various (17) visual recognition tasks. We further show that these visual recognition tasks can be categorically ordered based on their similarity to the source task such that a correlation between the performance of tasks and their similarity to the source task w.r.t. the proposed factors is observed.

Place, publisher, year, edition, pages
IEEE Computer Society Digital Library, 2016
National Category
Computer Vision and Robotics (Autonomous Systems)
Research subject
Computer Science
Identifiers
urn:nbn:se:kth:diva-177033 (URN)10.1109/TPAMI.2015.2500224 (DOI)000381432700006 ()2-s2.0-84981266620 (Scopus ID)
Note

QC 20161208

Available from: 2015-11-13 Created: 2015-11-13 Last updated: 2018-01-10Bibliographically approved
Azizpour, H. (2016). Visual Representations and Models: From Latent SVM to Deep Learning. (Doctoral dissertation). Stockholm, Sweden: KTH Royal Institute of Technology
Open this publication in new window or tab >>Visual Representations and Models: From Latent SVM to Deep Learning
2016 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

Two important components of a visual recognition system are representation and model. Both involves the selection and learning of the features that are indicative for recognition and discarding those features that are uninformative. This thesis, in its general form, proposes different techniques within the frameworks of two learning systems for representation and modeling. Namely, latent support vector machines (latent SVMs) and deep learning.

First, we propose various approaches to group the positive samples into clusters of visually similar instances. Given a fixed representation, the sampled space of the positive distribution is usually structured. The proposed clustering techniques include a novel similarity measure based on exemplar learning, an approach for using additional annotation, and augmenting latent SVM to automatically find clusters whose members can be reliably distinguished from background class. 

In another effort, a strongly supervised DPM is suggested to study how these models can benefit from privileged information. The extra information comes in the form of semantic parts annotation (i.e. their presence and location). And they are used to constrain DPMs latent variables during or prior to the optimization of the latent SVM. Its effectiveness is demonstrated on the task of animal detection.

Finally, we generalize the formulation of discriminative latent variable models, including DPMs, to incorporate new set of latent variables representing the structure or properties of negative samples. Thus, we term them as negative latent variables. We show this generalization affects state-of-the-art techniques and helps the visual recognition by explicitly searching for counter evidences of an object presence.

Following the resurgence of deep networks, in the last works of this thesis we have focused on deep learning in order to produce a generic representation for visual recognition. A Convolutional Network (ConvNet) is trained on a largely annotated image classification dataset called ImageNet with $\sim1.3$ million images. Then, the activations at each layer of the trained ConvNet can be treated as the representation of an input image. We show that such a representation is surprisingly effective for various recognition tasks, making it clearly superior to all the handcrafted features previously used in visual recognition (such as HOG in our first works on DPM). We further investigate the ways that one can improve this representation for a task in mind. We propose various factors involving before or after the training of the representation which can improve the efficacy of the ConvNet representation. These factors are analyzed on 16 datasets from various subfields of visual recognition.

Place, publisher, year, edition, pages
Stockholm, Sweden: KTH Royal Institute of Technology, 2016. p. 172
Series
TRITA-CSC-A, ISSN 1653-5723 ; 21
Keywords
Computer Vision, Machine Learning, Artificial Intelligence, Deep Learning, Learning Representation, Deformable Part Models, Discriminative Latent Variable Models, Convolutional Networks, Object Recognition, Object Detection
National Category
Electrical Engineering, Electronic Engineering, Information Engineering Computer Systems
Research subject
Computer Science
Identifiers
urn:nbn:se:kth:diva-192289 (URN)978-91-7729-110-7 (ISBN)
External cooperation:
Public defence
2016-09-27, Kollegiesalen, Brinellvägen 8, KTH-huset, våningsplan 4, KTH Campus, Stockholm, 15:26 (English)
Opponent
Supervisors
Note

QC 20160908

Available from: 2016-09-08 Created: 2016-09-08 Last updated: 2016-09-09Bibliographically approved
Azizpour, H., Razavian, A. S., Sullivan, J., Maki, A. & Carlsson, S. (2015). From Generic to Specific Deep Representations for Visual Recognition. In: Proceedings of CVPR 2015: . Paper presented at CVPRW DeepVision Workshop,June 11, 2015, Boston, MA, USA. IEEE conference proceedings
Open this publication in new window or tab >>From Generic to Specific Deep Representations for Visual Recognition
Show others...
2015 (English)In: Proceedings of CVPR 2015, IEEE conference proceedings, 2015Conference paper, Published paper (Refereed)
Abstract [en]

Evidence is mounting that ConvNets are the best representation learning method for recognition. In the common scenario, a ConvNet is trained on a large labeled dataset and the feed-forward units activation, at a certain layer of the network, is used as a generic representation of an input image. Recent studies have shown this form of representation to be astoundingly effective for a wide range of recognition tasks. This paper thoroughly investigates the transferability of such representations w.r.t. several factors. It includes parameters for training the network such as its architecture and parameters of feature extraction. We further show that different visual recognition tasks can be categorically ordered based on their distance from the source task. We then show interesting results indicating a clear correlation between the performance of tasks and their distance from the source task conditioned on proposed factors. Furthermore, by optimizing these factors, we achieve stateof-the-art performances on 16 visual recognition tasks.

Place, publisher, year, edition, pages
IEEE conference proceedings, 2015
Series
IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, ISSN 2160-7508
National Category
Computer Vision and Robotics (Autonomous Systems)
Identifiers
urn:nbn:se:kth:diva-164527 (URN)10.1109/CVPRW.2015.7301270 (DOI)000378887900005 ()2-s2.0-84951960494 (Scopus ID)978-146736759-2 (ISBN)
Conference
CVPRW DeepVision Workshop,June 11, 2015, Boston, MA, USA
Note

QC 20150507

Available from: 2015-04-17 Created: 2015-04-17 Last updated: 2018-01-11Bibliographically approved
Sharif Razavian, A., Azizpour, H., Maki, A., Sullivan, J., Ek, C. H. & Carlsson, S. (2015). Persistent Evidence of Local Image Properties in Generic ConvNets. In: Paulsen, Rasmus R., Pedersen, Kim S. (Ed.), Image Analysis: 19th Scandinavian Conference, SCIA 2015, Copenhagen, Denmark, June 15-17, 2015. Proceedings. Paper presented at Scandinavian Conference on Image Analysis, Copenhagen, Denmark, 15-17 June, 2015 (pp. 249-262). Springer Publishing Company
Open this publication in new window or tab >>Persistent Evidence of Local Image Properties in Generic ConvNets
Show others...
2015 (English)In: Image Analysis: 19th Scandinavian Conference, SCIA 2015, Copenhagen, Denmark, June 15-17, 2015. Proceedings / [ed] Paulsen, Rasmus R., Pedersen, Kim S., Springer Publishing Company, 2015, p. 249-262Conference paper, Published paper (Refereed)
Abstract [en]

Supervised training of a convolutional network for object classification should make explicit any information related to the class of objects and disregard any auxiliary information associated with the capture of the image or thevariation within the object class. Does this happen in practice? Although this seems to pertain to the very final layers in the network, if we look at earlier layers we find that this is not the case. Surprisingly, strong spatial information is implicit. This paper addresses this, in particular, exploiting the image representation at the first fully connected layer,i.e. the global image descriptor which has been recently shown to be most effective in a range of visual recognition tasks. We empirically demonstrate evidences for the finding in the contexts of four different tasks: 2d landmark detection, 2d object keypoints prediction, estimation of the RGB values of input image, and recovery of semantic label of each pixel. We base our investigation on a simple framework with ridge rigression commonly across these tasks,and show results which all support our insight. Such spatial information can be used for computing correspondence of landmarks to a good accuracy, but should potentially be useful for improving the training of the convolutional nets for classification purposes.

Place, publisher, year, edition, pages
Springer Publishing Company, 2015
Series
Image Processing, Computer Vision, Pattern Recognition, and Graphics ; 9127
National Category
Computer Systems
Identifiers
urn:nbn:se:kth:diva-172140 (URN)10.1007/978-3-319-19665-7_21 (DOI)2-s2.0-84947982864 (Scopus ID)
Conference
Scandinavian Conference on Image Analysis, Copenhagen, Denmark, 15-17 June, 2015
Note

Qc 20150828

Available from: 2015-08-13 Created: 2015-08-13 Last updated: 2016-12-09Bibliographically approved
Organisations
Identifiers
ORCID iD: ORCID iD iconorcid.org/0000-0001-5211-6388

Search in DiVA

Show all publications