Deep neural networks (DNNs) are known to be vulnerable to adversarial images, while their robustness in text classification is rarely studied. Several lines of text attack methods have been proposed in the literature, including character-level, word-level, and sentence-level attacks. However, it is still a challenge to minimize the number of word changes necessary to induce misclassification, while simultaneously ensuring lexical correctness, syntactic soundness, and semantic similarity. In this paper, we propose a Bigram and Unigram based adaptive Semantic Preservation Optimization (BU-SPO) method to examine the vulnerability of deep models. Our method has four major merits. Firstly, we propose to attack text documents not only at the unigram word level but also at the bigram level which better keeps semantics and avoids producing meaningless outputs. Secondly, we propose a hybrid method to replace the input words with options among both their synonyms candidates and sememe candidates, which greatly enriches the potential substitutions compared to only using synonyms. Thirdly, we design an optimization algorithm, i.e., Semantic Preservation Optimization (SPO), to determine the priority of word replacements, aiming to reduce the modification cost. Finally, we further improve the SPO with a semantic Filter (named SPOF) to find the adversarial example with the highest semantic similarity. We evaluate the effectiveness of our BU-SPO and BU-SPOF on IMDB, AG's News, and Yahoo! Answers text datasets by attacking four popular DNNs models. Results show that our methods achieve the highest attack success rates and semantics rates by changing the smallest number of words compared with existing methods.

Automatic Speech Recognition (ASR) is an active field of research due to its huge number of applications and the proliferation of interfaces or computing devices that can support speech processing. But the bulk of applications is based on well-resourced languages that overshadow under-resourced ones. Yet ASR represents an undeniable mean to promote such languages, especially when design human-to-human or human-to-machine systems involving illiterate people. An approach to design an ASR system targeting under-resourced languages is to start with a limited vocabulary. ASR using a limited vocabulary is a subset of the speech recognition problem that focuses on the recognition of a small number of words or sentences. This paper aims to provide a comprehensive view of mechanisms behind ASR systems as well as techniques, tools, projects, recent contributions, and possibly future directions in ASR using a limited vocabulary. This work consequently provides a way to go when designing ASR system using limited vocabulary. Although an emphasis is put on limited vocabulary, most of the tools and techniques reported in this survey applied to ASR systems in general.

The Deep Learning in Machine Vision market research provides detailed market development prospects, a market volume and value overview, and popular business trends. This research examined several elements of the demand for Deep Learning in Machine Vision. This study report goes into great detail about the many factors that have contributed to the Deep Learning in Machine Vision market's growth. A detailed analysis of international technology breakthroughs and developments is also included in Deep Learning in Machine Vision market research. Based on volume, performance, and valuation, the Deep Learning in Machine Vision industry analysis predicts the precise market share.

One of the several obstacles in the widespread use of AI systems is the lack of requirements of interpretability that can enable a layperson to ensure the safe and reliable behavior of such systems. We extend the analysis of an agent assessment module that lets an AI system execute high-level instruction sequences in simulators and answer the user queries about its execution of sequences of actions. We show that such a primitive query-response capability is sufficient to efficiently derive a user-interpretable causal model of the system in stationary, fully observable, and deterministic settings. We also introduce dynamic causal decision networks (DCDNs) that capture the causal structure of STRIPS-like domains. A comparative analysis of different classes of queries is also presented in terms of the computational requirements needed to answer them and the efforts required to evaluate their responses to learn the correct model.

COVID-19 continues to cause a significant impact on public health. To minimize this impact, policy makers undertake containment measures that however, when carried out disproportionately to the actual threat, as a result if errorneous threat assessment, cause undesirable long-term socio-economic complications. In addition, macro-level or national level decision making fails to consider the localized sensitivities in small regions. Hence, the need arises for region-wise threat assessments that provide insights on the behaviour of COVID-19 through time, enabled through accurate forecasts. In this study, a forecasting solution is proposed, to predict daily new cases of COVID-19 in regions small enough where containment measures could be locally implemented, by targeting three main shortcomings that exist in literature; the unreliability of existing data caused by inconsistent testing patterns in smaller regions, weak deploy-ability of forecasting models towards predicting cases in previously unseen regions, and model training biases caused by the imbalanced nature of data in COVID-19 epi-curves. Hence, the contributions of this study are three-fold; an optimized smoothing technique to smoothen less deterministic epi-curves based on epidemiological dynamics of that region, a Long-Short-Term-Memory (LSTM) based forecasting model trained using data from select regions to create a representative and diverse training set that maximizes deploy-ability in regions with lack of historical data, and an adaptive loss function whilst training to mitigate the data imbalances seen in epi-curves. The proposed smoothing technique, the generalized training strategy and the adaptive loss function largely increased the overall accuracy of the forecast, which enables efficient containment measures at a more localized micro-level.

Photojournalist documents the reality women face in Afghanistan on'America Reports' An Oklahoma mother of 11 flew to Qatar earlier this month to help rescue 10 members of Afghanistan's all-girls robotics team, and is hoping to save more as the Taliban takes power in Kabul. Allyson Reneau, a 60-year old-Harvard graduate with a Masters degree in international relations and U.S. space policy, took it upon herself to try and save members of the Afghan Girls Robotic Team, according to NBC. She flew into Qatar on Aug. 9 after making a "Hail Mary" call to a former roommate at the U.S. Embassy there to help get the girls from the advancing Taliban, known for their oppressive treatment of women. Reneau had been in contact with the team -- made of girls ages 16 to 18 -- since 2019, when she worked on the board of directors for Explore Mars and met the girls when they attended the organization's annual Humans to Mars conference. The team was hailed in Western media as the future of the war-ravaged country, as well as a shining example of how women's rights had improved after the U.S. invaded following 9/11.

More than 90 policy groups from the US and around the world signed an open letter urging Apple to drop its plan to have Apple devices scan photos for child sexual abuse material (CSAM). This story originally appeared on Ars Technica, a trusted source for technology news, tech policy analysis, reviews, and more. Ars is owned by WIRED's parent company, Condé Nast. "The undersigned organizations committed to civil rights, human rights, and digital rights around the world are writing to urge Apple to abandon the plans it announced on 5 August 2021 to build surveillance capabilities into iPhones, iPads, and other Apple products," the letter to Apple CEO Tim Cook said. "Though these capabilities are intended to protect children and to reduce the spread of child sexual abuse material (CSAM), we are concerned that they will be used to censor protected speech, threaten the privacy and security of people around the world, and have disastrous consequences for many children." The Center for Democracy and Technology (CDT) announced the letter, with CDT Security and Surveillance Project codirector Sharon Bradford Franklin saying, "We can expect governments will take advantage of the surveillance capability Apple is building into iPhones, iPads, and computers.

Michael Williams' wife pleaded with him to remember their fishing trips with the grandchildren, how he used to braid her hair, anything to jar him back to his world outside the concrete walls of Cook County Jail. His three daily calls to her had become a lifeline, but when they dwindled to two, then one, then only a few a week, the 65-year-old Williams felt he couldn't go on. He made plans to take his life with a stash of pills he had stockpiled in his dormitory. Williams was jailed last August, accused of killing a young man from the neighborhood who asked him for a ride during a night of unrest over police brutality in May. But the key evidence against Williams didn't come from an eyewitness or an informant; it came from a clip of noiseless security video showing a car driving through an intersection, and a loud bang picked up by a network of surveillance microphones. Prosecutors said technology powered by a secret algorithm that analyzed noises detected by the sensors indicated Williams shot and killed the man. "I kept trying to figure out, how can they get away with using the technology like that against me?" said Williams, speaking publicly for the first time about his ordeal. Williams sat behind bars for nearly a year before a judge dismissed the case against him last month at the request of prosecutors, who said they had insufficient evidence.

We introduce a new domain expert mixture (DEMix) layer that enables conditioning a language model (LM) on the domain of the input text. A DEMix layer is a collection of expert feedforward networks, each specialized to a domain, that makes the LM modular: experts can be mixed, added or removed after initial training. Extensive experiments with autoregressive transformer LMs (up to 1.3B parameters) show that DEMix layers reduce test-time perplexity, increase training efficiency, and enable rapid adaptation with little overhead. We show that mixing experts during inference, using a parameter-free weighted ensemble, allows the model to better generalize to heterogeneous or unseen domains. We also show that experts can be added to iteratively incorporate new domains without forgetting older ones, and that experts can be removed to restrict access to unwanted domains, without additional training. Overall, these results demonstrate benefits of explicitly conditioning on textual domains during language modeling.

Retail giant Shoprite on Wednesday revealed that it is testing an automated Checkers concept store with no cashiers, or till points." Checkers Rush is a "no queues, no checkout, no waiting" concept store, it said. "Using advanced AI camera technology to identify the products being taken off the shelves, Checkers Rush bills users' bank cards upon exit." The store is revealed in a promotional video for Shoprite X, the company's new digital innovation unit. The shop is available to staff at ShopriteX offices near the company's home office, above Checkers Hyper Brackenfell in Cape Town.