In multi-temporal SAR interferometry (MT-InSAR), persistent scatterer (PS) pixels are used to estimate geophysical parameters, essentially deformation. Conventionally, PS pixels are selected on the basis of the estimated noise present in the spatially uncorrelated phase component along with look-angle error in a temporal interferometric stack. In this study, two deep learning architectures, namely convolutional neural network for interferometric semantic segmentation (CNN-ISS) and convolutional long short term memory network for interferometric semantic segmentation (CLSTM-ISS), based on learning spatial and spatio-temporal behaviour respectively, were proposed for selection of PS pixels. These networks were trained to relate the interferometric phase history to its classification into phase stable (PS) and phase unstable (non-PS) measurement pixels using ~10,000 real world interferometric images of different study sites containing man-made objects, forests, vegetation, uncropped land, water bodies, and areas affected by lengthening, foreshortening, layover and shadowing. The networks were trained using training labels obtained from the Stanford method for Persistent Scatterer Interferometry (StaMPS) algorithm. However, pixel selection results, when compared to a combination of R-index and a classified image of the test dataset, reveal that CLSTM-ISS estimates improved the classification of PS and non-PS pixels compared to those of StaMPS and CNN-ISS. The predicted results show that CLSTM-ISS reached an accuracy of 93.50%, higher than that of CNN-ISS (89.21%). CLSTM-ISS also improved the density of reliable PS pixels compared to StaMPS and CNN-ISS and outperformed StaMPS and other conventional MT-InSAR methods in terms of computational efficiency.

The accurate automatic segmentation of gliomas and its intra-tumoral structures is important not only for treatment planning but also for follow-up evaluations. Several methods based on 2D and 3D Deep Neural Networks (DNN) have been developed to segment brain tumors and to classify different categories of tumors from different MRI modalities. However, these networks are often black-box models and do not provide any evidence regarding the process they take to perform this task. Increasing transparency and interpretability of such deep learning techniques are necessary for the complete integration of such methods into medical practice. In this paper, we explore various techniques to explain the functional organization of brain tumor segmentation models and to extract visualizations of internal concepts to understand how these networks achieve highly accurate tumor segmentations. We use the BraTS 2018 dataset to train three different networks with standard architectures and outline similarities and differences in the process that these networks take to segment brain tumors. We show that brain tumor segmentation networks learn certain human-understandable disentangled concepts on a filter level. We also show that they take a top-down or hierarchical approach to localizing the different parts of the tumor. We then extract visualizations of some internal feature maps and also provide a measure of uncertainty with regards to the outputs of the models to give additional qualitative evidence about the predictions of these networks. We believe that the emergence of such human-understandable organization and concepts might aid in the acceptance and integration of such methods in medical diagnosis.

Reinforcement learning (RL) is an effective approach to learn an optimal dialog policy for task-oriented visual dialog systems. A common practice is to apply RL on a neural sequence-to-sequence (seq2seq) framework with the action space being the output vocabulary in the decoder. However, it is difficult to design a reward function that can achieve a balance between learning an effective policy and generating a natural dialog response. This paper proposes a novel framework that alternatively trains a RL policy for image guessing and a supervised seq2seq model to improve dialog generation quality. We evaluate our framework on the GuessWhich task and the framework achieves the state-of-the-art performance in both task completion and dialog quality.

This paper focuses on the task of generating long structured sentences with explicit discourse markers, by proposing a new task Sentence Transfer and a novel model architecture TransSent. Previous works on text generation fused semantic and structure information in one mixed hidden representation. However, the structure was difficult to maintain properly when the generated sentence became longer. In this work, we explicitly separate the modeling process of semantic information and structure information. Intuitively, humans produce long sentences by directly connecting discourses with discourse markers like and, but, etc. We thus define a new task called Sentence Transfer. This task represents a long sentence as (head discourse, discourse marker, tail discourse) and aims at tail discourse generation based on head discourse and discourse marker. Then, by connecting original head discourse and generated tail discourse with a discourse marker, we generate a long structured sentence. We also propose a model architecture called TransSent, which models relations between two discourses by interpreting them as transferring from one discourse to the other in the embedding space. Experiment results show that our model achieves better performance in automatic evaluations, and can generate structured sentences with high quality. The datasets can be accessed by https://github.com/1024er/TransSent dataset.

Robustness of Deep Reinforcement Learning (DRL) algorithms towards adversarial attacks in real world applications such as those deployed in cyber-physical systems (CPS) are of increasing concern. Numerous studies have investigated the mechanisms of attacks on the RL agent's state space. Nonetheless, attacks on the RL agent's action space (AS) (corresponding to actuators in engineering systems) are equally perverse; such attacks are relatively less studied in the ML literature. In this work, we first frame the problem as an optimization problem of minimizing the cumulative reward of an RL agent with decoupled constraints as the budget of attack. We propose a white-box Myopic Action Space (MAS) attack algorithm that distributes the attacks across the action space dimensions. Next, we reformulate the optimization problem above with the same objective function, but with a temporally coupled constraint on the attack budget to take into account the approximated dynamics of the agent. This leads to the white-box Look-ahead Action Space (LAS) attack algorithm that distributes the attacks across the action and temporal dimensions. Our results shows that using the same amount of resources, the LAS attack deteriorates the agent's performance significantly more than the MAS attack. This reveals the possibility that with limited resource, an adversary can utilize the agent's dynamics to malevolently craft attacks that causes the agent to fail. Additionally, we leverage these attack strategies as a possible tool to gain insights on the potential vulnerabilities of DRL agents.

The rapid advancement of artificial intelligence (AI) is changing our lives in many ways. One application domain is data science. New techniques in automating the creation of AI, known as AutoAI or AutoML, aim to automate the work practices of data scientists. AutoAI systems are capable of autonomously ingesting and pre-processing data, engineering new features, and creating and scoring models based on a target objectives (e.g. accuracy or run-time efficiency). Though not yet widely adopted, we are interested in understanding how AutoAI will impact the practice of data science. We conducted interviews with 20 data scientists who work at a large, multinational technology company and practice data science in various business settings. Our goal is to understand their current work practices and how these practices might change with AutoAI. Reactions were mixed: while informants expressed concerns about the trend of automating their jobs, they also strongly felt it was inevitable. Despite these concerns, they remained optimistic about their future job security due to a view that the future of data science work will be a collaboration between humans and AI systems, in which both automation and human expertise are indispensable.

In this work, we aim to utilize prior knowledge encoded as logical rules to improve the performance of deep models. We propose a logic graph embedding network that projects d-DNNF formulae (and assignments) onto a manifold via an augmented Graph Convolutional Network (GCN). To generate semantically-faithful embeddings, we propose techniques to recognize node heterogeneity, and semantic regularization that incorporate structural constraints into the embedding. Experiments show that our approach improves the performance of models trained to perform model-checking and visual relation prediction.

Welcome to TNW Basics, a collection of tips, guides, and advice on how to easily get the most out of your gadgets, apps, and other stuff. This is also a part of our "Beginner's guide to AI," featuring articles on algorithms, neural networks, computer vision, natural language processing, and artificial general intelligence. The AI we use everyday in our phones, cameras, and smart devices usually falls into the category of deep learning. We've previously covered algorithms and artificial neural networks – concepts surrounding deep learning – but this time we'll take a look at how deep learning systems actually learn. Deep learning, to put it simply, is a method by which a machine can extract information from data by sending it through different layers of abstraction.

A synchronous input where all objects are presented simultaneously to a computer (left), versus an asynchronous input where objects are presented with temporal order... view more Machine learning, introduced 70 years ago, is based on evidence of the dynamics of learning in our brain. Using the speed of modern computers and large data sets, deep learning algorithms have recently produced results comparable to those of human experts in various applicable fields, but with different characteristics that are distant from current knowledge of learning in neuroscience. Using advanced experiments on neuronal cultures and large scale simulations, a group of scientists at Bar-Ilan University in Israel has demonstrated a new type of ultrafast artifical intelligence algorithms -- based on the very slow brain dynamics -- which outperform learning rates achieved to date by state-of-the-art learning algorithms. In an article published today in the journal Scientific Reports, the researchers rebuild the bridge between neuroscience and advanced artificial intelligence algorithms that has been left virtually useless for almost 70 years. "The current scientific and technological viewpoint is that neurobiology and machine learning are two distinct disciplines that advanced independently," said the study's lead author, Prof. Ido Kanter, of Bar-Ilan University's Department of Physics and Gonda (Goldschmied) Multidisciplinary Brain Research Center.

A version of this story appeared in STAT Health Tech, our weekly newsletter about how tech is transforming health care and the life sciences. Sign up here to receive it in your inbox. You're looking at two versions of the same video of a moment in a single cell, captured under a powerful microscope. The red and yellow structures are mitochondria, and the inset magnified in the bottom left hand corner in each view captures a mitochondrion dividing. The view on the left shows the raw data as it came off the microscope; you might think of it like a social media influencer's first take, before any filters have been applied to get that Instagram-ready look.