Human motion prediction is a fundamental part of many human-robot applications. Despite the recent progress in human motion prediction, most studies simplify the problem by predicting the human motion relative to a fixed joint and/or only limit their model to predict one possible future motion. While due to the complex nature of human motion, a single output cannot reflect all the possible actions one can do. Also, for any robotics application, we need the full human motion including the user trajectory not a 3d pose relative to the hip joint. In this paper, we try to address these two issues by proposing a transformer-based generative model for forecasting multiple diverse human motions. Our model generates \textit{N} future possible motion by querying a history of human motion. Our model first predicts the pose of the body relative to the hip joint. Then the \textit{Hip Prediction Module} predicts the trajectory of the hip movement for each predicted pose frame. To emphasize on the diverse future motions we introduce a similarity loss that penalizes the pairwise sample distance. We show that our system outperforms the state-of-the-art in human motion prediction while it can predict diverse multi-motion future trajectories with hip movements

Question answering (QA) has demonstrated impressive progress in answering questions from customized domains. Nevertheless, domain adaptation remains one of the most elusive challenges for QA systems, especially when QA systems are trained in a source domain but deployed in a different target domain. In this work, we investigate the potential benefits of question classification for QA domain adaptation. We propose a novel framework: Question Classification for Question Answering (QC4QA). Specifically, a question classifier is adopted to assign question classes to both the source and target data. Then, we perform joint training in a self-supervised fashion via pseudo-labeling. For optimization, inter-domain discrepancy between the source and target domain is reduced via maximum mean discrepancy (MMD) distance. We additionally minimize intra-class discrepancy among QA samples of the same question class for fine-grained adaptation performance. To the best of our knowledge, this is the first work in QA domain adaptation to leverage question classification with self-supervised adaptation. We demonstrate the effectiveness of the proposed QC4QA with consistent improvements against the state-of-the-art baselines on multiple datasets.

Ten years ago, if you bandied around the word Artificial Intelligence in a Kenyan newsroom, you would be branded an illusionist, today it is the buzzword. AI, machine learning and data processing are fast becoming essential in any newsroom that hopes to survive the digital revolution avalanche. In fact, job titles such as data scientist, data analyst, content strategist, content engineer and impact editors are fast gaining currency in media houses deemed digital journalism early adopters. Today, it is almost impossible to distinguish between an article authored by a machine and that written by a journalist, AI has enabled editors to figure out their audiences needs to granular levels and offer them articles that "speak to their souls". If the trend of automation of journalism catches on in the continent fast enough, in a few years it would be difficult to tell if the news anchor calling you "mpendwa msikilizaji" is a robot or human.

The hard-and software that catalysed rapid developments in machine learning In late 2002 and early 2003, the release of the Radeon 9700 and GeForce FX video cards introduced a fully programmable graphics pipeline, extending and later replacing the existing fixed function pipelines. Unlike the fixed function pipeline, which allowed the user to only supply input matrices and parameters to built-in operations, the programmable pipeline introduced the execution of user-written shader programs on the GPU [Contributors, 2015]. This fundamental change allowed programmers and researchers to exploit the intrinsic parallelism of GPUs 2 years before Intel would introduce its first dual-core CPU. Within months of the availability of this new hardware and the accompanying APIs, researchers implemented linear algebra methods on GPUs and introduced programming frameworks to use GPUs for generalpurpose computations [Thompson et al., 2002, Krüger and Westermann, 2003]. This rapid development marked the dawn of general-purpose computing on graphics processing units (GPGPU). In a presentation at ICS '08, Harris presented the successes of GPGPU by highlighting a speed-up in molecular docking, N-body simulations, HD video stream transcoding, or image processing--applications in machine learning were not discussed. However, just one year later, the introduction of GPUs as general-purpose processors catalysed the deep learning explosion of the early 2010s by allowing deep learning algorithms pioneered by Alexey Ivakhnenko in 1971 to be run within practical time on widely available consumer hardware when Rajat et al. showed that GPUs outperform CPUs by an order of magnitude in large-scale deep unsupervised learning tasks [Ivakhnenko, 1971, Raina et al., 2009]. Hardware and energy requirements increase in machine learning research In 2010, Ciresan et al. [2010] introduced a multi-layer perceptron (MLP) with up to 12.11 million free parameters where forward and backward propagation were implemented on a GPU using NVIDIA's proprietary CUDA API introduced by Harris at ICS '08 two

To address three significant challenges of massive wireless communications including propagation loss, long-distance transmission, and channel fading, we aim at establishing the hybrid antenna array with bowtie elements in a compact size for beamforming applications. In this work we rigorously demonstrate that bowtie elements allow for a significant improvement in the beamforming performance of the hybrid antenna array compared to not only other available antenna arrays, but also its geometrical counterpart with dipole elements. We have achieved a greater than 15 dB increase in SINR values, a greater than 20% improvement in the antenna efficiency, a significant enhancement in the DoA estimation, and 20 increments in the directivity for the hybrid antenna array with bowtie elements, compared to its geometrical counterpart, by performing a three-class classification LS-SVM (Least-Squares Support Vector Machine) optimization method. The proposed hybrid antenna array has shown a 3D uniform directivity, which is accompanied by its superior performance in the 3D uniform beam-scanning capability. The directivities remain almost constant at 40.83 dBi with the variation of angle θ, and 41.21 dBi with the variation of angle φ. The unrivaled functionality and performance of the hybrid antenna array with bowtie elements makes it a potential candidate for beamforming applications in massive wireless communications. ECENT advances and innovative solutions have been extensively developed for deploying smart array antennas with a highly-directive radiation pattern in order to overcome the long-distance challenges and minimize interference signals, as well as enhance the spherical coverage and capacity in massive wireless communication channels. In this sense, in addition to performing optimization methods, the geometrical configuration, inter-element spacing, excitation phase, and amplitude of the individual array elements of each antenna array have been designed and synthesized for controlling its radiation patterns.

The ACII Affective Vocal Bursts (A-VB) competition introduces a new topic in affective computing, which is understanding emotional expression using the non-verbal sound of humans. We are familiar with emotion recognition via verbal vocal or facial expression. However, the vocal bursts such as laughs, cries, and signs, are not exploited even though they are very informative for behavior analysis. The A-VB competition comprises four tasks that explore non-verbal information in different spaces. This technical report describes the method and the result of SclabCNU Team for the tasks of the challenge. We achieved promising results compared to the baseline model provided by the organizers.

Smart face mask could detect coronavirus on other people's breath

LONDON, Sept. 27, 2022 (GLOBE NEWSWIRE) -- According to The Business Research Company's research report on the high energy lasers market, the emergence of artificial intelligence is gaining popularity among the high energy lasers industry trends. Many companies operating in the market are focused on developing AI-based products to get a competitive advantage. For instance, in April 2022, the US Navy successfully tested the Layered Laser Defense (LLD), a laser weapon designed and developed by Lockheed Martin, a US-based aerospace, arms, defense, information security, and technology company. This is the Layered Laser Defense (LLD). It can use a high-power laser to counter unmanned aerial systems and fast-attack boats, as well as track inbound air threats, support combat identification, and conduct battle damage assessments of engaged targets.

Anyone who has followed global news events of late will have noticed the devastating floods that have engulfed pretty much every corner of the world, from the U.S. and Europe, to Africa, Australia and Asia, where India and Pakistan have been hit by some of their worst floods in recent memory. By pretty much all accounts, such climate change-driven disasters are only going to get worse. And while there are varying opinions on what -- if anything -- we can do to avert such catastrophes in the future, some companies are looking at ways to plan for this new reality, and at least go some way toward mitigating the impact of flooding. One of these companies is 7Analytics, a Norwegian startup founded back in 2020 by a team of data scientists and geologists to reduce the risks of flooding for construction and energy infrastructure companies. With its first product, FloodCube, 7Analytics serves customers with AI and advanced machine learning techniques to calculate current surface water and where it's flowing today (the "runoff"), then models how that will look in the future with increased rainfall.

Understanding how the brain encodes external stimuli and how these stimuli can be decoded from the measured brain activities are long-standing and challenging questions in neuroscience. In this paper, we focus on reconstructing the complex image stimuli from fMRI (functional magnetic resonance imaging) signals. Unlike previous works that reconstruct images with single objects or simple shapes, our work aims to reconstruct image stimuli that are rich in semantics, closer to everyday scenes, and can reveal more perspectives. However, data scarcity of fMRI datasets is the main obstacle to applying state-of-the-art deep learning models to this problem. We find that incorporating an additional text modality is beneficial for the reconstruction problem compared to directly translating brain signals to images. Therefore, the modalities involved in our method are: (i) voxel-level fMRI signals, (ii) observed images that trigger the brain signals, and (iii) textual description of the images. To further address data scarcity, we leverage an aligned vision-language latent space pre-trained on massive datasets. Instead of training models from scratch to find a latent space shared by the three modalities, we encode fMRI signals into this pre-aligned latent space. Then, conditioned on embeddings in this space, we reconstruct images with a generative model. The reconstructed images from our pipeline balance both naturalness and fidelity: they are photo-realistic and capture the ground truth image contents well.