A US university's claim it can use facial recognition to "predict criminality" has renewed debate over racial bias in technology. Harrisburg University researchers said their software "can predict if someone is a criminal, based solely on a picture of their face". The software "is intended to help law enforcement prevent crime", it said. But 1,700 academics have signed an open letter demanding the research remains unpublished. One Harrisburg research member, a former police officer, wrote: "Identifying the criminality of [a] person from their facial image will enable a significant advantage for law-enforcement agencies and other intelligence agencies to prevent crime from occurring."

The news: US Democratic lawmakers have introduced a bill that would ban the use of facial recognition technology by federal law enforcement agencies. Specifically, it would make it illegal for any federal agency or official to "acquire, possess, access, or use" biometric surveillance technology in the US. It would also require state and local law enforcement to bring in similar bans in order to receive federal funding. The Facial Recognition and Biometric Technology Moratorium Act was introduced by Senators Ed Markey of Massachusetts and Jeff Merkley of Oregon and Representatives Pramila Jayapal of Washington and Ayanna Pressley of Massachusetts. Seize the moment: The proposed law has arrived at a point when the police use of facial recognition technology is coming under increased scrutiny amid protests after the killing of George Floyd in late May.

A fleet of steam-powered robots could one day be used to explore the moons of Jupiter and Saturn by'taking giant leaps across their frozen landscapes'. NASA Jet Propulsion Laboratory researchers are developing small football-sized spherical robots equipped with steam-powered thrusters for future exploration. If they ever move beyond the initial concept stage, the robots would let scientists explore icy moons such as Europa and Enceladus orbiting Jupiter and Saturn. These moons are thought to harbour salty subsurface oceans of liquid water, but very little is known about their surface - making it a potentially challenging terrain for a traditional moon rover, but easy for a leaping robotic ball powered by steam. The bots, called SPARROW, would run on steam from ice that was collected by mining the surfaces of the moons they explore - rather than'dirty' rocket fuel.

Lawmakers have proposed legislation that would indefinitely ban the use of facial recognition technology by law enforcement nationwide. The new bill comes after months of public concerns surrounding facial recognition's implications for data privacy, government surveillance and racial bias. The Facial Recognition and Biometric Technology Moratorium Act was proposed Thursday by Sens. Ed Markey (D-MA) and Jeff Merkley (D-OR), and Reps. While various cities have banned government use of the technology (with Boston this week becoming the tenth U.S. city to do so), the bill would be the first temporary ban on facial recognition technology ever enacted nationwide. The newly proposed bill would "prohibit biometric surveillance by the Federal Government without explicit statutory authorization and to withhold certain Federal public safety grants from State and local governments that engage in biometric surveillance."

Geneva – More than 50 countries, including Japan, South Korea and the European Union member states, have agreed common regulations for vehicles that can take over some driving functions, including having a mandatory black box, the U.N. announced Thursday. The binding rules on Automated Lane Keeping Systems (ALKS) will come into force in January 2021. The measures were adopted by the United Nations Economic Commission for Europe (UNECE) World Forum for Harmonization of Vehicle Regulations, which brings together 53 countries, not just in Europe but also in Africa and Asia. "This is the first binding international regulation on so-called'Level 3' vehicle automation," UNECE said in a statement. "The new regulation therefore marks an important step towards the wider deployment of automated vehicles to help realize a vision of safer, more sustainable mobility for all."

Successful algorithms have been developed for computing Nash equilibrium in a variety of finite game classes. However, solving continuous games---in which the pure strategy space is (potentially uncountably) infinite---is far more challenging. Nonetheless, many real-world domains have continuous action spaces, e.g., where actions refer to an amount of time, money, or other resource that is naturally modeled as being real-valued as opposed to integral. We present a new algorithm for computing Nash equilibrium strategies in continuous games. In addition to two-player zero-sum games, our algorithm also applies to multiplayer games and games of imperfect information. We experiment with our algorithm on a continuous imperfect-information Blotto game, in which two players distribute resources over multiple battlefields. Blotto games have frequently been used to model national security scenarios and have also been applied to electoral competition and auction theory. Experiments show that our algorithm is able to quickly compute close approximations of Nash equilibrium strategies for this game.

Our goal is to enable machine learning systems to be trained interactively. This requires models that perform well and train quickly, without large amounts of hand-labeled data. We take a step forward in this direction by borrowing from weak supervision (WS), wherein models can be trained with noisy sources of signal instead of hand-labeled data. But WS relies on training downstream deep networks to extrapolate to unseen data points, which can take hours or days. Pre-trained embeddings can remove this requirement. We do not use the embeddings as features as in transfer learning (TL), which requires fine-tuning for high performance, but instead use them to define a distance function on the data and extend WS source votes to nearby points. Theoretically, we provide a series of results studying how performance scales with changes in source coverage, source accuracy, and the Lipschitzness of label distributions in the embedding space, and compare this rate to standard WS without extension and TL without fine-tuning. On six benchmark NLP and video tasks, our method outperforms WS without extension by 4.1 points, TL without fine-tuning by 12.8 points, and traditionally-supervised deep networks by 13.1 points, and comes within 0.7 points of state-of-the-art weakly-supervised deep networks--all while training in less than half a second.

We introduce MARGE, a pre-trained sequence-to-sequence model learned with an unsupervised multi-lingual multi-document paraphrasing objective. MARGE provides an alternative to the dominant masked language modeling paradigm, where we self-supervise the reconstruction of target text by retrieving a set of related texts (in many languages) and conditioning on them to maximize the likelihood of generating the original. We show it is possible to jointly learn to do retrieval and reconstruction, given only a random initialization. The objective noisily captures aspects of paraphrase, translation, multi-document summarization, and information retrieval, allowing for strong zero-shot performance on several tasks. For example, with no additional task-specific training we achieve BLEU scores of up to 35.8 for document translation. We further show that fine-tuning gives strong performance on a range of discriminative and generative tasks in many languages, making MARGE the most generally applicable pre-training method to date.

Research in adversarial machine learning has shown how the performance of machine learning models can be seriously compromised by injecting even a small fraction of poisoning points into the training data. While the effects on model accuracy of such poisoning attacks have been widely studied, their potential effects on other model performance metrics remain to be evaluated. In this work, we introduce an optimization framework for poisoning attacks against algorithmic fairness, and develop a gradient-based poisoning attack aimed at introducing classification disparities among different groups in the data. We empirically show that our attack is effective not only in the white-box setting, in which the attacker has full access to the target model, but also in a more challenging black-box scenario in which the attacks are optimized against a substitute model and then transferred to the target model. We believe that our findings pave the way towards the definition of an entirely novel set of adversarial attacks targeting algorithmic fairness in different scenarios, and that investigating such vulnerabilities will help design more robust algorithms and countermeasures in the future.

An open question in autonomous driving is how best to use simulation to validate the safety of autonomous vehicles. Existing techniques rely on simulated rollouts, which can be inefficient for finding rare failure events, while other techniques are designed to only discover a single failure. In this work, we present a new safety validation approach that attempts to estimate the distribution over failures of an autonomous policy using approximate dynamic programming. Knowledge of this distribution allows for the efficient discovery of many failure examples. To address the problem of scalability, we decompose complex driving scenarios into subproblems consisting of only the ego vehicle and one other vehicle. These subproblems can be solved with approximate dynamic programming and their solutions are recombined to approximate the solution to the full scenario. We apply our approach to a simple two-vehicle scenario to demonstrate the technique as well as a more complex five-vehicle scenario to demonstrate scalability. In both experiments, we observed an increase in the number of failures discovered compared to baseline approaches.