Models, code, and papers for "Rob van de Loo":

Epithelium segmentation using deep learning in H&E-stained prostate specimens with immunohistochemistry as reference standard

Aug 17, 2018
Wouter Bulten, Péter Bándi, Jeffrey Hoven, Rob van de Loo, Johannes Lotz, Nick Weiss, Jeroen van der Laak, Bram van Ginneken, Christina Hulsbergen-van de Kaa, Geert Litjens

Prostate cancer (PCa) is graded by pathologists by examining the architectural pattern of cancerous epithelial tissue on hematoxylin and eosin (H&E) stained slides. Given the importance of gland morphology, automatically differentiating between glandular epithelial tissue and other tissues is an important prerequisite for the development of automated methods for detecting PCa. We propose a new method, using deep learning, for automatically segmenting epithelial tissue in digitized prostatectomy slides. We employed immunohistochemistry (IHC) to render the ground truth less subjective and more precise compared to manual outlining on H&E slides, especially in areas with high-grade and poorly differentiated PCa. Our dataset consisted of 102 tissue blocks, including both low and high grade PCa. From each block a single new section was cut, stained with H&E, scanned, restained using P63 and CK8/18 to highlight the epithelial structure, and scanned again. The H&E slides were co-registered to the IHC slides. On a subset of the IHC slides we applied color deconvolution, corrected stain errors manually, and trained a U-Net to perform segmentation of epithelial structures. Whole-slide segmentation masks generated by the IHC U-Net were used to train a second U-Net on H&E. Our system makes precise cell-level segmentations and segments both intact glands as well as individual (tumor) epithelial cells. We achieved an F1-score of 0.895 on a hold-out test set and 0.827 on an external reference set from a different center. We envision this segmentation as being the first part of a fully automated prostate cancer detection and grading pipeline.

  Click for Model/Code and Paper
Comparison of Different Methods for Tissue Segmentation in Histopathological Whole-Slide Images

Apr 03, 2017
Péter Bándi, Rob van de Loo, Milad Intezar, Daan Geijs, Francesco Ciompi, Bram van Ginneken, Jeroen van der Laak, Geert Litjens

Tissue segmentation is an important pre-requisite for efficient and accurate diagnostics in digital pathology. However, it is well known that whole-slide scanners can fail in detecting all tissue regions, for example due to the tissue type, or due to weak staining because their tissue detection algorithms are not robust enough. In this paper, we introduce two different convolutional neural network architectures for whole slide image segmentation to accurately identify the tissue sections. We also compare the algorithms to a published traditional method. We collected 54 whole slide images with differing stains and tissue types from three laboratories to validate our algorithms. We show that while the two methods do not differ significantly they outperform their traditional counterpart (Jaccard index of 0.937 and 0.929 vs. 0.870, p < 0.01).

* Accepted for poster presentation at the IEEE International Symposium on Biomedical Imaging (ISBI) 2017 

  Click for Model/Code and Paper
Whole-Slide Mitosis Detection in H&E Breast Histology Using PHH3 as a Reference to Train Distilled Stain-Invariant Convolutional Networks

Aug 17, 2018
David Tellez, Maschenka Balkenhol, Irene Otte-Holler, Rob van de Loo, Rob Vogels, Peter Bult, Carla Wauters, Willem Vreuls, Suzanne Mol, Nico Karssemeijer, Geert Litjens, Jeroen van der Laak, Francesco Ciompi

Manual counting of mitotic tumor cells in tissue sections constitutes one of the strongest prognostic markers for breast cancer. This procedure, however, is time-consuming and error-prone. We developed a method to automatically detect mitotic figures in breast cancer tissue sections based on convolutional neural networks (CNNs). Application of CNNs to hematoxylin and eosin (H&E) stained histological tissue sections is hampered by: (1) noisy and expensive reference standards established by pathologists, (2) lack of generalization due to staining variation across laboratories, and (3) high computational requirements needed to process gigapixel whole-slide images (WSIs). In this paper, we present a method to train and evaluate CNNs to specifically solve these issues in the context of mitosis detection in breast cancer WSIs. First, by combining image analysis of mitotic activity in phosphohistone-H3 (PHH3) restained slides and registration, we built a reference standard for mitosis detection in entire H&E WSIs requiring minimal manual annotation effort. Second, we designed a data augmentation strategy that creates diverse and realistic H&E stain variations by modifying the hematoxylin and eosin color channels directly. Using it during training combined with network ensembling resulted in a stain invariant mitosis detector. Third, we applied knowledge distillation to reduce the computational requirements of the mitosis detection ensemble with a negligible loss of performance. The system was trained in a single-center cohort and evaluated in an independent multicenter cohort from The Cancer Genome Atlas on the three tasks of the Tumor Proliferation Assessment Challenge (TUPAC). We obtained a performance within the top-3 best methods for most of the tasks of the challenge.

* Accepted to appear in IEEE Transactions on Medical Imaging 

  Click for Model/Code and Paper
A weakly supervised sequence tagging and grammar induction approach to semantic frame slot filling

Jun 15, 2019
Janneke van de Loo, Guy De Pauw, Walter Daelemans

This paper describes continuing work on semantic frame slot filling for a command and control task using a weakly-supervised approach. We investigate the advantages of using retraining techniques that take the output of a hierarchical hidden markov model as input to two inductive approaches: (1) discriminative sequence labelers based on conditional random fields and memory-based learning and (2) probabilistic context-free grammar induction. Experimental results show that this setup can significantly improve F-scores without the need for additional information sources. Furthermore, qualitative analysis shows that the weakly supervised technique is able to automatically induce an easily interpretable and syntactically appropriate grammar for the domain and task at hand.

  Click for Model/Code and Paper
Model-Based Action Exploration for Learning Dynamic Motion Skills

Apr 12, 2018
Glen Berseth, Michiel van de Panne

Deep reinforcement learning has achieved great strides in solving challenging motion control tasks. Recently, there has been significant work on methods for exploiting the data gathered during training, but there has been less work on how to best generate the data to learn from. For continuous action domains, the most common method for generating exploratory actions involves sampling from a Gaussian distribution centred around the mean action output by a policy. Although these methods can be quite capable, they do not scale well with the dimensionality of the action space, and can be dangerous to apply on hardware. We consider learning a forward dynamics model to predict the result, ($x_{t+1}$), of taking a particular action, ($u$), given a specific observation of the state, ($x_{t}$). With this model we perform internal look-ahead predictions of outcomes and seek actions we believe have a reasonable chance of success. This method alters the exploratory action space, thereby increasing learning speed and enables higher quality solutions to difficult problems, such as robotic locomotion and juggling.

* 7 pages, 7 figures, conference paper 

  Click for Model/Code and Paper
MoNoise: Modeling Noise Using a Modular Normalization System

Oct 10, 2017
Rob van der Goot, Gertjan van Noord

We propose MoNoise: a normalization model focused on generalizability and efficiency, it aims at being easily reusable and adaptable. Normalization is the task of translating texts from a non- canonical domain to a more canonical domain, in our case: from social media data to standard language. Our proposed model is based on a modular candidate generation in which each module is responsible for a different type of normalization action. The most important generation modules are a spelling correction system and a word embeddings module. Depending on the definition of the normalization task, a static lookup list can be crucial for performance. We train a random forest classifier to rank the candidates, which generalizes well to all different types of normaliza- tion actions. Most features for the ranking originate from the generation modules; besides these features, N-gram features prove to be an important source of information. We show that MoNoise beats the state-of-the-art on different normalization benchmarks for English and Dutch, which all define the task of normalization slightly different.

* Source code: 

  Click for Model/Code and Paper
A Three-Player GAN: Generating Hard Samples To Improve Classification Networks

Mar 08, 2019
Simon Vandenhende, Bert De Brabandere, Davy Neven, Luc Van Gool

We propose a Three-Player Generative Adversarial Network to improve classification networks. In addition to the game played between the discriminator and generator, a competition is introduced between the generator and the classifier. The generator's objective is to synthesize samples that are both realistic and hard to label for the classifier. Even though we make no assumptions on the type of augmentations to learn, we find that the model is able to synthesize realistically looking examples that are hard for the classification model. Furthermore, the classifier becomes more robust when trained on these difficult samples. The method is evaluated on a public dataset for traffic sign recognition.

* Accepted for oral presentation at MVA2019 

  Click for Model/Code and Paper
Neural Networks for Information Retrieval

Jul 13, 2017
Tom Kenter, Alexey Borisov, Christophe Van Gysel, Mostafa Dehghani, Maarten de Rijke, Bhaskar Mitra

Machine learning plays a role in many aspects of modern IR systems, and deep learning is applied in all of them. The fast pace of modern-day research has given rise to many different approaches for many different IR problems. The amount of information available can be overwhelming both for junior students and for experienced researchers looking for new research topics and directions. Additionally, it is interesting to see what key insights into IR problems the new technologies are able to give us. The aim of this full-day tutorial is to give a clear overview of current tried-and-trusted neural methods in IR and how they benefit IR research. It covers key architectures, as well as the most promising future directions.

* Overview of full-day tutorial at SIGIR 2017 

  Click for Model/Code and Paper
Explaining First Impressions: Modeling, Recognizing, and Explaining Apparent Personality from Videos

Oct 15, 2018
Hugo Jair Escalante, Heysem Kaya, Albert Ali Salah, Sergio Escalera, Yagmur Gucluturk, Umut Guclu, Xavier Baro, Isabelle Guyon, Julio Jacques Junior, Meysam Madadi, Stephane Ayache, Evelyne Viegas, Furkan Gurpinar, Achmadnoer Sukma Wicaksana, Cynthia C. S. Liem, Marcel A. J. van Gerven, Rob van Lier

Explainability and interpretability are two critical aspects of decision support systems. Within computer vision, they are critical in certain tasks related to human behavior analysis such as in health care applications. Despite their importance, it is only recently that researchers are starting to explore these aspects. This paper provides an introduction to explainability and interpretability in the context of computer vision with an emphasis on looking at people tasks. Specifically, we review and study those mechanisms in the context of first impressions analysis. To the best of our knowledge, this is the first effort in this direction. Additionally, we describe a challenge we organized on explainability in first impressions analysis from video. We analyze in detail the newly introduced data set, the evaluation protocol, and summarize the results of the challenge. Finally, derived from our study, we outline research opportunities that we foresee will be decisive in the near future for the development of the explainable computer vision field.

* Preprint submitted to IJCV 

  Click for Model/Code and Paper
You Only Look Twice: Rapid Multi-Scale Object Detection In Satellite Imagery

May 24, 2018
Adam Van Etten

Detection of small objects in large swaths of imagery is one of the primary problems in satellite imagery analytics. While object detection in ground-based imagery has benefited from research into new deep learning approaches, transitioning such technology to overhead imagery is nontrivial. Among the challenges is the sheer number of pixels and geographic extent per image: a single DigitalGlobe satellite image encompasses >64 km2 and over 250 million pixels. Another challenge is that objects of interest are minuscule (often only ~10 pixels in extent), which complicates traditional computer vision techniques. To address these issues, we propose a pipeline (You Only Look Twice, or YOLT) that evaluates satellite images of arbitrary size at a rate of >0.5 km2/s. The proposed approach can rapidly detect objects of vastly different scales with relatively little training data over multiple sensors. We evaluate large test images at native resolution, and yield scores of F1 > 0.8 for vehicle localization. We further explore resolution and object size requirements by systematically testing the pipeline at decreasing resolution, and conclude that objects only ~5 pixels in size can still be localized with high confidence. Code is available at

* 8 pages, 14 figures, 3 tables 

  Click for Model/Code and Paper
From Algorithmic Black Boxes to Adaptive White Boxes: Declarative Decision-Theoretic Ethical Programs as Codes of Ethics

Nov 16, 2017
Martijn van Otterlo

Ethics of algorithms is an emerging topic in various disciplines such as social science, law, and philosophy, but also artificial intelligence (AI). The value alignment problem expresses the challenge of (machine) learning values that are, in some way, aligned with human requirements or values. In this paper I argue for looking at how humans have formalized and communicated values, in professional codes of ethics, and for exploring declarative decision-theoretic ethical programs (DDTEP) to formalize codes of ethics. This renders machine ethical reasoning and decision-making, as well as learning, more transparent and hopefully more accountable. The paper includes proof-of-concept examples of known toy dilemmas and gatekeeping domains such as archives and libraries.

* 7 pages, 1 figure, submitted 

  Click for Model/Code and Paper
A theory of consciousness: computation, algorithm, and neurobiological realization

Aug 01, 2018
J. H. van Hateren

The most enigmatic aspect of consciousness is the fact that it is felt, as a subjective sensation. This particular aspect is explained by the theory proposed here. The theory encompasses both the computation that is presumably involved and the way in which that computation may be realized in the brain's neurobiology. It is assumed that the brain makes an internal estimate of an individual's own evolutionary fitness, which can be shown to produce an irreducible, distinct cause. Communicating components of the fitness estimate (either for external or internal use) requires inverting them. Such inversion can be performed by the thalamocortical feedback loop in the mammalian brain, if that loop is operating in a switched, dual-stage mode. A first (nonconscious) stage produces forward estimates, whereas the second (conscious) stage inverts those estimates. It is argued that inversion produces irreducible, distinct, and spatially localized causes, which are plausibly sensed as the feeling of consciousness.

* revision, 22 pages, 12 figures 

  Click for Model/Code and Paper
Active causation and the origin of meaning

May 25, 2014
J. H. van Hateren

Purpose and meaning are necessary concepts for understanding mind and culture, but appear to be absent from the physical world and are not part of the explanatory framework of the natural sciences. Understanding how meaning (in the broad sense of the term) could arise from a physical world has proven to be a tough problem. The basic scheme of Darwinian evolution produces adaptations that only represent apparent ("as if") goals and meaning. Here I use evolutionary models to show that a slight, evolvable extension of the basic scheme is sufficient to produce genuine goals. The extension, targeted modulation of mutation rate, is known to be generally present in biological cells, and gives rise to two phenomena that are absent from the non-living world: intrinsic meaning and the ability to initiate goal-directed chains of causation (active causation). The extended scheme accomplishes this by utilizing randomness modulated by a feedback loop that is itself regulated by evolutionary pressure. The mechanism can be extended to behavioural variability as well, and thus shows how freedom of behaviour is possible. A further extension to communication suggests that the active exchange of intrinsic meaning between organisms may be the origin of consciousness, which in combination with active causation can provide a physical basis for the phenomenon of free will.

* Biological Cybernetics 109, 33-46 (2015) 
* revised and extended 

  Click for Model/Code and Paper
Conflict-Directed Backjumping Revisited

Jun 01, 2011
X. Chen, P. van Beek

In recent years, many improvements to backtracking algorithms for solving constraint satisfaction problems have been proposed. The techniques for improving backtracking algorithms can be conveniently classified as look-ahead schemes and look-back schemes. Unfortunately, look-ahead and look-back schemes are not entirely orthogonal as it has been observed empirically that the enhancement of look-ahead techniques is sometimes counterproductive to the effects of look-back techniques. In this paper, we focus on the relationship between the two most important look-ahead techniques---using a variable ordering heuristic and maintaining a level of local consistency during the backtracking search---and the look-back technique of conflict-directed backjumping (CBJ). We show that there exists a "perfect" dynamic variable ordering such that CBJ becomes redundant. We also show theoretically that as the level of local consistency that is maintained in the backtracking search is increased, the less that backjumping will be an improvement. Our theoretical results partially explain why a backtracking algorithm doing more in the look-ahead phase cannot benefit more from the backjumping look-back scheme. Finally, we show empirically that adding CBJ to a backtracking algorithm that maintains generalized arc consistency (GAC), an algorithm that we refer to as GAC-CBJ, can still provide orders of magnitude speedups. Our empirical results contrast with Bessiere and Regin's conclusion (1996) that CBJ is useless to an algorithm that maintains arc consistency.

* Journal Of Artificial Intelligence Research, Volume 14, pages 53-81, 2001 

  Click for Model/Code and Paper
Episodic Multi-armed Bandits

Mar 11, 2018
Cem Tekin, Mihaela van der Schaar

We introduce a new class of reinforcement learning methods referred to as {\em episodic multi-armed bandits} (eMAB). In eMAB the learner proceeds in {\em episodes}, each composed of several {\em steps}, in which it chooses an action and observes a feedback signal. Moreover, in each step, it can take a special action, called the $stop$ action, that ends the current episode. After the $stop$ action is taken, the learner collects a terminal reward, and observes the costs and terminal rewards associated with each step of the episode. The goal of the learner is to maximize its cumulative gain (i.e., the terminal reward minus costs) over all episodes by learning to choose the best sequence of actions based on the feedback. First, we define an {\em oracle} benchmark, which sequentially selects the actions that maximize the expected immediate gain. Then, we propose our online learning algorithm, named {\em FeedBack Adaptive Learning} (FeedBAL), and prove that its regret with respect to the benchmark is bounded with high probability and increases logarithmically in expectation. Moreover, the regret only has polynomial dependence on the number of steps, actions and states. eMAB can be used to model applications that involve humans in the loop, ranging from personalized medical screening to personalized web-based education, where sequences of actions are taken in each episode, and optimal behavior requires adapting the chosen actions based on the feedback.

  Click for Model/Code and Paper
Reinforcement Learning with Non-uniform State Representations for Adaptive Search

Jun 15, 2019
Sandeep Manjanna, Herke van Hoof, Gregory Dudek

Efficient spatial exploration is a key aspect of search and rescue. In this paper, we present a search algorithm that generates efficient trajectories that optimize the rate at which probability mass is covered by a searcher. This should allow an autonomous vehicle find one or more lost targets as rapidly as possible. We do this by performing non-uniform sampling of the search region. The path generated minimizes the expected time to locate the missing target by visiting high probability regions using non-myopic path generation based on reinforcement learning. We model the target probability distribution using a classic mixture of Gaussians model with means and mixture coefficients tuned according to the location and time of sightings of the lost target. Key features of our search algorithm are the ability to employ a very general non-deterministic action model and the ability to generate action plans for any new probability distribution using the parameters learned on other similar looking distributions. One of the key contributions of this paper is the use of non-uniform state aggregation for policy search in the context of robotics.


  Click for Model/Code and Paper
Learning with Value-Ramp

Apr 23, 2017
Tom J. Ameloot, Jan Van den Bussche

We study a learning principle based on the intuition of forming ramps. The agent tries to follow an increasing sequence of values until the agent meets a peak of reward. The resulting Value-Ramp algorithm is natural, easy to configure, and has a robust implementation with natural numbers.

* Version 2: fixed notation in definition of transition + clarified a sentence in the Introduction 

  Click for Model/Code and Paper
Enaction-Based Artificial Intelligence: Toward Coevolution with Humans in the Loop

Feb 26, 2014
Pierre De Loor, Kristen Manach, Jacques Tisseau

This article deals with the links between the enaction paradigm and artificial intelligence. Enaction is considered a metaphor for artificial intelligence, as a number of the notions which it deals with are deemed incompatible with the phenomenal field of the virtual. After explaining this stance, we shall review previous works regarding this issue in terms of artifical life and robotics. We shall focus on the lack of recognition of co-evolution at the heart of these approaches. We propose to explicitly integrate the evolution of the environment into our approach in order to refine the ontogenesis of the artificial system, and to compare it with the enaction paradigm. The growing complexity of the ontogenetic mechanisms to be activated can therefore be compensated by an interactive guidance system emanating from the environment. This proposition does not however resolve that of the relevance of the meaning created by the machine (sense-making). Such reflections lead us to integrate human interaction into this environment in order to construct relevant meaning in terms of participative artificial intelligence. This raises a number of questions with regards to setting up an enactive interaction. The article concludes by exploring a number of issues, thereby enabling us to associate current approaches with the principles of morphogenesis, guidance, the phenomenology of interactions and the use of minimal enactive interfaces in setting up experiments which will deal with the problem of artificial intelligence in a variety of enaction-based ways.

* Minds and Machine, num 19, pp 319-343, 2009 

  Click for Model/Code and Paper
A Conflict-Based Path-Generation Heuristic for Evacuation Planning

Sep 10, 2013
Victor Pillac, Pascal Van Henetenryck, Caroline Even

Evacuation planning and scheduling is a critical aspect of disaster management and national security applications. This paper proposes a conflict-based path-generation approach for evacuation planning. Its key idea is to generate evacuation routes lazily for evacuated areas and to optimize the evacuation over these routes in a master problem. Each new path is generated to remedy conflicts in the evacuation and adds new columns and a new row in the master problem. The algorithm is applied to massive flood scenarios in the Hawkesbury-Nepean river (West Sydney, Australia) which require evacuating in the order of 70,000 persons. The proposed approach reduces the number of variables from 4,500,000 in a Mixed Integer Programming (MIP) formulation to 30,000 in the case study. With this approach, realistic evacuations scenarios can be solved near-optimally in real time, supporting both evacuation planning in strategic, tactical, and operational environments.

* Technical report 

  Click for Model/Code and Paper
An FPGA-based Massively Parallel Neuromorphic Cortex Simulator

Mar 08, 2018
Runchun Wang, Chetan Singh Thakur, Andre van Schaik

This paper presents a massively parallel and scalable neuromorphic cortex simulator designed for simulating large and structurally connected spiking neural networks, such as complex models of various areas of the cortex. The main novelty of this work is the abstraction of a neuromorphic architecture into clusters represented by minicolumns and hypercolumns, analogously to the fundamental structural units observed in neurobiology. Without this approach, simulating large-scale fully connected networks needs prohibitively large memory to store look-up tables for point-to-point connections. Instead, we use a novel architecture, based on the structural connectivity in the neocortex, such that all the required parameters and connections can be stored in on-chip memory. The cortex simulator can be easily reconfigured for simulating different neural networks without any change in hardware structure by programming the memory. A hierarchical communication scheme allows one neuron to have a fan-out of up to 200k neurons. As a proof-of-concept, an implementation on one Altera Stratix V FPGA was able to simulate 20 million to 2.6 billion leaky-integrate-and-fire (LIF) neurons in real time. We verified the system by emulating a simplified auditory cortex (with 100 million neurons). This cortex simulator achieved a low power dissipation of 1.62 {\mu}W per neuron. With the advent of commercially available FPGA boards, our system offers an accessible and scalable tool for the design, real-time simulation, and analysis of large-scale spiking neural networks.

* 18 pages 

  Click for Model/Code and Paper