Apple sold 2.4 million smart speakers in the U.S. market during first quarter of 2021, according to the market research firm Omdia, beating Google's sales by 100,000 units during the same period. The firm estimates the new Apple HomePod mini accounted for 91 percent of Apple's U.S. smart speaker sales during that time. Apple's share of smart speaker shipments in the U.S. market reached 17.8 percent during the first quarter, a 9 percent increase year over year. But Omdia estimates there were 75 million smart speakers based on Google Assistant on the market at the end of 2020, compared to just 10 million based on Apple's Siri. And Amazon dominates both of those competitors, with a current installed base of 141 million smart speakers using its Alexa voice assistant.

The UneeQ, based in the United States and New Zealand, published a video of its artificial intelligence project Digital Einstein that has the father of relativity theory chat with a fictional version of his human Sofia. Users of UneeQ technology will be able to chat with the iconic Nobel Prize in Physics, who will answer their questions. The idea of this long-term project is to teach and accompany people who feel lonely, especially seeing the effects of quarantines around the world due to the COVID-19 pandemic. The company said in a statement that "Digital Einstein, among other digital humans, can communicate with people in a more natural way: using conversation, human expressions and emotional responses to provide the best daily interactions that we hope will make a difference in people's lives ".

It's "misleading and counterproductive" to block the use of machine-learning algorithms in the justice system on the grounds that some of them may be subject to racial bias, according to a forthcoming study in the American Criminal Law Review. The use of artificial intelligence by judges, prosecutors, police and other justice authorities remains "the best means to overcome the pervasive bias and discrimination that exists in all parts of the deeply flawed criminal justice system," said the study. Algorithmic systems are used in a variety of ways in the U.S. justice system in practices ranging from identifying and predicting crime "hot spots" to real-time surveillance. More than 60 kinds of risk assessment tools are currently in use by court systems around the country, usually to weigh whether individuals should be held in detention before trial or can be released on their own recognizance. The risk assessment tools, which assign weights to data points such as previous arrests and the age of the offender, have come under fire from activists, judges, prosecutors, and some criminologists who say they are susceptible to bias themselves.

This paper presents a technique to interpret and visualize intermediate layers in CNNs trained on raw speech data in an unsupervised manner. We show that averaging over feature maps after ReLU activation in each convolutional layer yields interpretable time-series data. The proposed technique enables acoustic analysis of intermediate convolutional layers. To uncover how meaningful representation in speech gets encoded in intermediate layers of CNNs, we manipulate individual latent variables to marginal levels outside of the training range. We train and probe internal representations on two models -- a bare WaveGAN architecture and a ciwGAN extension which forces the Generator to output informative data and results in emergence of linguistically meaningful representations. Interpretation and visualization is performed for three basic acoustic properties of speech: periodic vibration (corresponding to vowels), aperiodic noise vibration (corresponding to fricatives), and silence (corresponding to stops). We also argue that the proposed technique allows acoustic analysis of intermediate layers that parallels the acoustic analysis of human speech data: we can extract F0, intensity, duration, formants, and other acoustic properties from intermediate layers in order to test where and how CNNs encode various types of information. The models are trained on two speech processes with different degrees of complexity: a simple presence of [s] and a computationally complex presence of reduplication (copied material). Observing the causal effect between interpolation and the resulting changes in intermediate layers can reveal how individual variables get transformed into spikes in activation in intermediate layers. Using the proposed technique, we can analyze how linguistically meaningful units in speech get encoded in different convolutional layers.

Object detection is one of the fundamental objectives in Applied Computer Vision. In some of the applications, object detection becomes very challenging such as in the case of satellite image processing. Satellite image processing has remained the focus of researchers in domains of Precision Agriculture, Climate Change, Disaster Management, etc. Therefore, object detection in satellite imagery is one of the most researched problems in this domain. This paper focuses on aircraft detection. in satellite imagery using deep learning techniques. In this paper, we used YOLO deep learning framework for aircraft detection. This method uses satellite images collected by different sources as learning for the model to perform detection. Object detection in satellite images is mostly complex because objects have many variations, types, poses, sizes, complex and dense background. YOLO has some limitations for small size objects (less than$\sim$32 pixels per object), therefore we upsample the prediction grid to reduce the coarseness of the model and to accurately detect the densely clustered objects. The improved model shows good accuracy and performance on different unknown images having small, rotating, and dense objects to meet the requirements in real-time.

As the vocabulary size of modern word-based language models becomes ever larger, many sampling-based training criteria are proposed and investigated. The essence of these sampling methods is that the softmax-related traversal over the entire vocabulary can be simplified, giving speedups compared to the baseline. A problem we notice about the current landscape of such sampling methods is the lack of a systematic comparison and some myths about preferring one over another. In this work, we consider Monte Carlo sampling, importance sampling, a novel method we call compensated partial summation, and noise contrastive estimation. Linking back to the three traditional criteria, namely mean squared error, binary cross-entropy, and cross-entropy, we derive the theoretical solutions to the training problems. Contrary to some common belief, we show that all these sampling methods can perform equally well, as long as we correct for the intended class posterior probabilities. Experimental results in language modeling and automatic speech recognition on Switchboard and LibriSpeech support our claim, with all sampling-based methods showing similar perplexities and word error rates while giving the expected speedups.

The investment of time and resources for better strategies and methodologies to tackle a potential pandemic is key to deal with potential outbreaks of new variants or other viruses in the future. In this work, we recreated the scene of a year ago, 2020, when the pandemic erupted across the world for the fifty countries with more COVID-19 cases reported. We performed some experiments in which we compare state-of-the-art machine learning algorithms, such as LSTM, against online incremental machine learning algorithms to adapt them to the daily changes in the spread of the disease and predict future COVID-19 cases. To compare the methods, we performed three experiments: In the first one, we trained the models using only data from the country we predicted. In the second one, we use data from all fifty countries to train and predict each of them. In the first and second experiment, we used a static hold-out approach for all methods. In the third experiment, we trained the incremental methods sequentially, using a prequential evaluation. This scheme is not suitable for most state-of-the-art machine learning algorithms because they need to be retrained from scratch for every batch of predictions, causing a computational burden. Results show that incremental methods are a promising approach to adapt to changes of the disease over time; they are always up to date with the last state of the data distribution, and they have a significantly lower computational cost than other techniques such as LSTMs.

Building a dialogue system that can communicate naturally with humans is a challenging yet interesting problem of agent-based computing. The rapid growth in this area is usually hindered by the long-standing problem of data scarcity as these systems are expected to learn syntax, grammar, decision making, and reasoning from insufficient amounts of task-specific dataset. The recently introduced pre-trained language models have the potential to address the issue of data scarcity and bring considerable advantages by generating contextualized word embeddings. These models are considered counterpart of ImageNet in NLP and have demonstrated to capture different facets of language such as hierarchical relations, long-term dependency, and sentiment. In this short survey paper, we discuss the recent progress made in the field of pre-trained language models. We also deliberate that how the strengths of these language models can be leveraged in designing more engaging and more eloquent conversational agents. This paper, therefore, intends to establish whether these pre-trained models can overcome the challenges pertinent to dialogue systems, and how their architecture could be exploited in order to overcome these challenges. Open challenges in the field of dialogue systems have also been deliberated.

Knowledge Graph (KG) reasoning that predicts missing facts for incomplete KGs has been widely explored. However, reasoning over Temporal KG (TKG) that predicts facts in the future is still far from resolved. The key to predict future facts is to thoroughly understand the historical facts. A TKG is actually a sequence of KGs corresponding to different timestamps, where all concurrent facts in each KG exhibit structural dependencies and temporally adjacent facts carry informative sequential patterns. To capture these properties effectively and efficiently, we propose a novel Recurrent Evolution network based on Graph Convolution Network (GCN), called RE-GCN, which learns the evolutional representations of entities and relations at each timestamp by modeling the KG sequence recurrently. Specifically, for the evolution unit, a relation-aware GCN is leveraged to capture the structural dependencies within the KG at each timestamp. In order to capture the sequential patterns of all facts in parallel, the historical KG sequence is modeled auto-regressively by the gate recurrent components. Moreover, the static properties of entities such as entity types, are also incorporated via a static graph constraint component to obtain better entity representations. Fact prediction at future timestamps can then be realized based on the evolutional entity and relation representations. Extensive experiments demonstrate that the RE-GCN model obtains substantial performance and efficiency improvement for the temporal reasoning tasks on six benchmark datasets. Especially, it achieves up to 11.46\% improvement in MRR for entity prediction with up to 82 times speedup comparing to the state-of-the-art baseline.

In 2018, in late October, a distinctly odd painting appeared at the fine art auction house Christe's. At a distance, the painting looks like a 19th-century portrait of an austere gentleman dressed in black. Contained in a gilt frame, the portly gentleman appears middle-aged; his white-collar insinuates that he is a man of the church. The painting seems unassuming, something expected at an auction house that sells billions of dollars of painting each year. However, upon closer inspection, things get a bit odd.