According to reports, the growth of the global OTT marketplace is accelerating much more because of the pandemic. Previously, the compound annual growth rate (CAGR) was 16%. Post-pandemic, this number is expected to reach 19% by 2026, the market value could reach US$438.5 billion. King of the OTT platforms, Netflix is well-known for providing high-quality content with endless options. Netflix is globally popular for its services, and the secret behind is technologies like AI and machine learning that serve more relevant and intuitive recommendations to audiences.

To develop faster computers, batteries with higher capacities, and lighter and stronger automobiles and airplanes, scientists are working to discover and synthesize new materials with high-performance properties. But the number of atom combinations that could compose new materials is nearly unlimited, which makes the discovery process time consuming and expensive. To speed up the process, scientists are using theoretical models and computers to explore all the possibilities and disregard materials that are not desirable. Ali Davariashtiyani, a PhD student working under the direction of Assistant Professor Sara Kadkhodaei in the Computational Materials Research Lab at UIC, has developed a data-driven deep-learning model to help researchers identify easily synthesizable materials. Their findings were recently published in the journal Communications Materials.

Innovation including artificial intelligence that prioritize audience preferences and by extension embrace new market dynamics will ensure the media industry survives for the next several decades to come. Innovations such as use of drones, AI and related have helped the media improve quality of and speed of releasing content to a level not imagine before. Media houses are using AI to maximize on audience segmentation and preferences, thus gradually we are seeing a stabilization of changes in ratings, trust and credibility in the media, as the industry finds footing in the fast -changing operating environment. Can they help in ensuring responsible and professional of media reporting of elections, especially within the context of compressed newsrooms? Can such be the panacea to misinformation and propaganda that is becoming a big threat to professional elections reporting and related safety challenges to journalists during the electioneering period?

In research funded by the firm (which is yet to be peer-reviewed), Prof George Oikonomou and his team compared mobility scores for cattle made by two human experts with those made by CattleEye. They found that the technology was roughly 80-90% in agreement with the two experts - in terms of judging which animals were lame.

Recommender systems typically suggest to users content similar to what they consumed in the past. If a user happens to be exposed to strongly polarized content, she might subsequently receive recommendations which may steer her towards more and more radicalized content, eventually being trapped in what we call a "radicalization pathway". In this paper, we study the problem of mitigating radicalization pathways using a graph-based approach. Specifically, we model the set of recommendations of a "what-to-watch-next" recommender as a d-regular directed graph where nodes correspond to content items, links to recommendations, and paths to possible user sessions. We measure the "segregation" score of a node representing radicalized content as the expected length of a random walk from that node to any node representing non-radicalized content. High segregation scores are associated to larger chances to get users trapped in radicalization pathways. Hence, we define the problem of reducing the prevalence of radicalization pathways by selecting a small number of edges to "rewire", so to minimize the maximum of segregation scores among all radicalized nodes, while maintaining the relevance of the recommendations. We prove that the problem of finding the optimal set of recommendations to rewire is NP-hard and NP-hard to approximate within any factor. Therefore, we turn our attention to heuristics, and propose an efficient yet effective greedy algorithm based on the absorbing random walk theory. Our experiments on real-world datasets in the context of video and news recommendations confirm the effectiveness of our proposal.

Audio classification is an important task of mapping audio samples into their corresponding labels. Recently, the transformer model with self-attention mechanisms has been adopted in this field. However, existing audio transformers require large GPU memories and long training time, meanwhile relying on pretrained vision models to achieve high performance, which limits the model's scalability in audio tasks. To combat these problems, we introduce HTS-AT: an audio transformer with a hierarchical structure to reduce the model size and training time. It is further combined with a token-semantic module to map final outputs into class featuremaps, thus enabling the model for the audio event detection (i.e. localization in time). We evaluate HTS-AT on three datasets of audio classification where it achieves new state-of-the-art (SOTA) results on AudioSet and ESC-50, and equals the SOTA on Speech Command V2. It also achieves better performance in event localization than the previous CNN-based models. Moreover, HTS-AT requires only 35% model parameters and 15% training time of the previous audio transformer. These results demonstrate the high performance and high efficiency of HTS-AT.

Generating music with deep neural networks has been an area of active research in recent years. While the quality of generated samples has been steadily increasing, most methods are only able to exert minimal control over the generated sequence, if any. We propose the self-supervised description-to-sequence task, which allows for fine-grained controllable generation on a global level. We do so by extracting high-level features about the target sequence and learning the conditional distribution of sequences given the corresponding high-level description in a sequence-to-sequence modelling setup. We train FIGARO (FIne-grained music Generation via Attention-based, RObust control) by applying description-to-sequence modelling to symbolic music. By combining learned high level features with domain knowledge, which acts as a strong inductive bias, the model achieves state-of-the-art results in controllable symbolic music generation and generalizes well beyond the training distribution.

Robot with violin is followed by cloned businessmen. For decades, scientists and tech visionaries have envisioned a day when computers become so powerful that they become smarter than the human race. There is no shortage of science fiction stories and movies about robot uprisings. We are very far from that scary scenario, but at the same time artificial intelligence (AI) is no longer sci-fi. Many applications of AI abound today in business, and it is even being used in creative professions.

While the pandemic slowed down the development of businesses and entire industries, it did not affect the ongoing development of AI-generated speech. According to analysts at Meticulous Research, the global voice technology market is growing at 17.2% annually. By 2025 its volume is expected to reach $26.8 billion. What makes voice synthesis such a rapidly developing niche, and what impact is that development having on speech-based applications today? Implementing speech-based applications helps businesses significantly improve customer experiences.

New York, US, Jan. 31, 2022 (GLOBE NEWSWIRE) -- Market Overview: According to a comprehensive research report by Market Research Future (MRFR), "AI as a Service Market information by Technology, by Vertical and Region – forecast to 2030" market size to reach USD 43.29 billion, growing at a compound annual growth rate of 25.8% by 2030. AIaaS Market Scope: The increasing expenditure to adopt AI and advances in technology for workflow optimization will offer robust opportunities for the AlaaS market over the forecast period. Besides, the increasing adoption of cloud-based solutions in different end user industries & increasing need for cognitive computing will also fuel market growth. Besides, the other factors adding market growth include the increasing use of social media platforms, increase in the number of start-ups, and increasing demand for artificial intelligence enabled SDK's and APIs. Market USP Exclusively Encompassed: AIaaS Market Drivers Growing Need for AI and Cognitive Computing to Boost Market Growth The growing need for artificial intelligence and cognitive computing & the large-scale use of cloud-based solutions for intelligent business applications will boost market growth over the forecast period.