Convolutional Neural Networks (CNNs) allowed improving the state-of-the-art for many vision applications. However, naive CNNs suffer from two serious issues: vulnerability to adversarial examples and making incorrect but confident predictions for out-distribution samples. In this paper, we draw a connection between these two issues of CNNs through over-generalization. We reveal an augmented CNN (an extra output class added) as a simple yet effective end-to-end approach has the capacity for controlling over-generalization. We demonstrate training an augmented CNN on only a properly selected natural out-distribution dataset and interpolated samples empowers it to classify a wide range of unseen out-distribution samples as dustbin. Meanwhile, its misclassification rates on a broad spectrum of well-known black-box adversaries drop drastically as it classifies a portion of adversaries as dustbin class (rejection option) while correctly classifies some of the remaining. However, such an augmented CNN is never trained with any types of adversaries. Finally, generation of white-box adversarial attacks using augmented CNNs can be harder as the attack algorithms have to avoid dustbin regions for generating actual adversaries.

Abstract--Deep neural networks for image classification typically consists of a convolutional feature extractor followed by a fully connected classifier network. The predicted and the ground truth labels are represented as one hot vectors. Such a representation assumes that all classes are equally dissimilar . However, classes have visual similarities and often form a hierarchy. Learning this latent hierarchy explicitly in the architecture could provide invaluable insights. We propose an alternate architecture to the classifier network called the Latent Hierarchy (LH) Classifier and an end to end learned Class2Str mapping which discovers a latent hierarchy of the classes. We show that for some of the best performing architectures on CIF AR and Imagenet datasets, the proposed replacement and training by LH classifier recovers the accuracy, with a fraction of the number of parameters in the classifier part. Compared to the previous work of HDCNN, which also learns a 2 level hierarchy, we are able to learn a hierarchy at an arbitrary number of levels as well as obtain an accuracy improvement on the Imagenet classification task over them. We also verify that many visually similar classes are grouped together, under the learnt hierarchy.

Nefarious actors on social media and other platforms often spread rumors and falsehoods through images whose metadata (e.g., captions) have been modified to provide visual substantiation of the rumor/falsehood. This type of modification is referred to as image repurposing, in which often an unmanipulated image is published along with incorrect or manipulated metadata to serve the actor's ulterior motives. We present the Multimodal Entity Image Repurposing (MEIR) dataset, a substantially challenging dataset over that which has been previously available to support research into image repurposing detection. The new dataset includes location, person, and organization manipulations on real-world data sourced from Flickr. We also present a novel, end-to-end, deep multimodal learning model for assessing the integrity of an image by combining information extracted from the image with related information from a knowledge base. The proposed method is compared against state-of-the-art techniques on existing datasets as well as MEIR, where it outperforms existing methods across the board, with AUC improvement up to 0.23.

Approaches to goal recognition have progressively relaxed the requirements about the amount of domain knowledge and available observations, yielding accurate and efficient algorithms capable of recognizing goals. However, to recognize goals in raw data, recent approaches require either human engineered domain knowledge, or samples of behavior that account for almost all actions being observed to infer possible goals. This is clearly too strong a requirement for real-world applications of goal recognition, and we develop an approach that leverages advances in recurrent neural networks to perform goal recognition as a classification task, using encoded plan traces for training. We empirically evaluate our approach against the state-of-the-art in goal recognition with image-based domains, and discuss under which conditions our approach is superior to previous ones.

Reinforcement learning methods have been used for learning dialogue policies from the experience of conversations. However, learning an effective dialogue policy frequently requires prohibitively many conversations. This is partly because of the sparse rewards in dialogues, and the relatively small number of successful dialogues in early learning phase. Hindsight experience replay (HER) enables an agent to learn from failure, but the vanilla HER is inapplicable to dialogue domains due to dialogue goals being implicit (c.f., explicit goals in manipulation tasks). In this work, we develop two complex HER methods providing different trade-offs between complexity and performance. Experiments were conducted using a realistic user simulator. Results suggest that our HER methods perform better than standard and prioritized experience replay methods (as applied to deep Q-networks) in learning rate, and that our two complex HER methods can be combined to produce the best performance.

Recently, network lasso has drawn many attentions due to its remarkable performance on simultaneous clustering and optimization. However, it usually suffers from the imperfect data (noise, missing values etc), and yields sub-optimal solutions. The reason is that it finds the similar instances according to their features directly, which is usually impacted by the imperfect data, and thus returns sub-optimal results. In this paper, we propose triangle lasso to avoid its disadvantage. Triangle lasso finds the similar instances according to their neighbours. If two instances have many common neighbours, they tend to become similar. Although some instances are profiled by the imperfect data, it is still able to find the similar counterparts. Furthermore, we develop an efficient algorithm based on Alternating Direction Method of Multipliers (ADMM) to obtain a moderately accurate solution. In addition, we present a dual method to obtain the accurate solution with the low additional time consumption. We demonstrate through extensive numerical experiments that triangle lasso is robust to the imperfect data. It usually yields a better performance than the state-of-the-art method when performing data analysis tasks in practical scenarios.

A decade ago, the prospect of a religion that worships Artificial Intelligence would have seemed absurd, a fringe delusion both socially unacceptable and technologically improbable. In the last several years, however, advances in machine learning, robotics, cognitive science, genetic editing, and other fields have given rise to the belief that the destiny of our species will be determined by technology--whether it saves us or destroys us. Although the machine-as-god theme has appeared in science fiction as far back as far back as Isaac Asimov's short stories "The Last Question" and "Reason," and more recently in films like The Matrix and iRobot, the divinization of AI is no longer merely a fancy of fiction. It has become a mainstream metaphor, as evidenced by the growing number of scientists who openly describe technological progress in religious terms, including Hans Peter Moravec, Allen Newell, Ray Kurzweil, and Hugo de Garis. But this drive to replace the old gods and old religions with the new ones of science and technology doesn't stop at metaphor.

While Europe ranks second in the global artificial intelligence start-up landscape, no single European country achieves critical mass on its own. Europe needs to consolidate its efforts in order to compete against China and the US.

Mobile transportation platform provider, Didi Chuxing (DiDi), has launched an initiative to fully understand the social impact on communities of artificial intelligence (AI). The company, which offers a range of app-based transportation options for 550 million users including taxi, carpooling and bike-share, made the announcement at its Tech Day in Beijing, where it also unveiled its latest developments in AI with industry partners, academia, students and engineers. At the event, DiDi's tech executives, led by Bob Zhang Bo, chief technology officer, shared details of breakthroughs in the company's popular Express Pool service, which aims to improve the user experience by using machine learning algorithms. Operating across more 100 cities in China, Express Pool now accounts for over 20 per cent of all trips in many core markets. It said the algorithms can help carpooling become the most effective mobility option for drivers to increase their income, for passengers to reduce costs, and for cities to enjoy a high level of efficiency in the transportation system.

Japan is planning to accept a greater number of caregivers from three Southeast Asian countries having bilateral free trade agreements with Tokyo as part of efforts to address a labor shortage in the country, sources familiar with the matter said Sunday. By easing some restrictions on the number of caregivers from Indonesia, the Philippines and Vietnam, the government will allow more caregivers with high Japanese language skills to work in Japan from next April, the sources said. Under the current terms with the three countries, Japan will accept up to 300 caregivers from each country a year. The government aims to treat such foreign workers with high language proficiency separately from the current quota of 300. The number of people in the three countries who want to work as caregivers in Japan has recently been increasing.