To market its new series about an evil artificial intelligence creation threatening society, Fox commissioned a trailer created by--an actual AI. The network partnered with digital agency Space150 to assemble a trailer for its thriller Next, debuting Tuesday, that was entirely written, edited and scored by AI. The tech-focused creative shop developed a machine learning algorithm that watched the entire series, and selected key themes, scenes and dialogue for what appears to be linear TV's first AI-created trailer. Last February, Space150 creative shop was behind the "Travis Bott," a fake Travis Scott song and video created entirely by AI trained on the rapper's music. That caught the attention of Fox executives, which had been looking for a unique way to market Next.

Speaker verification (SV) has recently attracted considerable research interest due to the growing popularity of virtual assistants. At the same time, there is an increasing requirement for an SV system: it should be robust to short speech segments, especially in noisy and reverberant environments. In this paper, we consider one more important requirement for practical applications: the system should be robust to an audio stream containing long non-speech segments, where a voice activity detection (VAD) is not applied. To meet these two requirements, we introduce feature pyramid module (FPM)-based multi-scale aggregation (MSA) and self-adaptive soft VAD (SAS-VAD). We present the FPM-based MSA to deal with short speech segments in noisy and reverberant environments. Also, we use the SAS-VAD to increase the robustness to long non-speech segments. To further improve the robustness to acoustic distortions (i.e., noise and reverberation), we apply a masking-based speech enhancement (SE) method. We combine SV, VAD, and SE models in a unified deep learning framework and jointly train the entire network in an end-to-end manner. To the best of our knowledge, this is the first work combining these three models in a deep learning framework. We conduct experiments on Korean indoor (KID) and VoxCeleb datasets, which are corrupted by noise and reverberation. The results show that the proposed method is effective for SV in the challenging conditions and performs better than the baseline i-vector and deep speaker embedding systems.

Recent advances in Knowledge Graph Embedding (KGE) allow for representing entities and relations in continuous vector spaces. Some traditional KGE models leveraging additional type information can improve the representation of entities which however totally rely on the explicit types or neglect the diverse type representations specific to various relations. Besides, none of the existing methods is capable of inferring all the relation patterns of symmetry, inversion and composition as well as the complex properties of 1-N, N-1 and N-N relations, simultaneously. To explore the type information for any KG, we develop a novel KGE framework with Automated Entity TypE Representation (AutoETER), which learns the latent type embedding of each entity by regarding each relation as a translation operation between the types of two entities with a relation-aware projection mechanism. Particularly, our designed automated type representation learning mechanism is a pluggable module which can be easily incorporated with any KGE model. Besides, our approach could model and infer all the relation patterns and complex relations. Experiments on four datasets demonstrate the superior performance of our model compared to state-of-the-art baselines on link prediction tasks, and the visualization of type clustering provides clearly the explanation of type embeddings and verifies the effectiveness of our model.

In a quiet aisle of a small supermarket in Tokyo, a robot dutifully goes about its work. It looks like a well-integrated autonomous mechanical worker, but that is something of an illusion. This robot doesn't have a mind of its own. Several miles away, a human worker is controlling its every movement remotely and watching via a virtual reality (VR) headset that provides a robot's eye view. This is the work of Japanese firm Telexistence, whose Model-T robot is designed to allow people to do physical labour in supermarkets and other locations from the comfort of their own homes.

AWS DeepComposer provides a creative and hands-on experience for learning generative AI and machine learning (ML). We recently launched the Edit melody feature, which allows you to add, remove, or edit specific notes, giving you full control of the pitch, length, and timing for each note. In this post, you can learn to use the Edit melody feature to collaborate with the autoregressive convolutional neural network (AR-CNN) algorithm and create interesting Bach-style compositions. Through human-AI collaboration, we can surpass what humans and AI systems can create independently. For example, you can seek inspiration from AI to create art or music outside their area of expertise or offload the more routine tasks, like creating variations on a melody, and focus on the more interesting and creative tasks.

To whet your appetite a bit more for Assassin's Creed Valhalla, Ubisoft wants to teach you about the history of Vikings. It made a podcast called Echoes of Valhalla and all five episodes are now available on Spotify. Ubisoft described the documentary series as "the first immersive audio historical documentary series in audio for Assassin's Creed." With the help of experts, comedians and "reconstructed scenes," it delves into various aspects of Viking life, such as shipbuilding, military strategy and the role of women. Although the podcast is a first for Assassin's Creed, Ubisoft has tried to educate players about the games' historical settings through other means.

Nest Audio costs $99.99 and comes in five colors. There hasn't been a new Google Home speaker since way back in 2016 and now Google is giving its OG smart speaker a major upgrade with the Nest Audio. We're still getting to know the Nest Audio, but here are some important specs to note: The speaker keeps the same fabric-wrapped look of previous Google speakers like the Nest Mini, but it takes on a more unique oval-shape and it's taller than many other smart speakers on the market. It doesn't come with any USB-C or auxiliary input ports, but unlike the Sonos One, it supports both WiFi and Bluetooth connection. As you'd expect in a modern smart speaker, the Nest can be paired with other Nest speakers for multi-room audio.

The old Google Home that looked like an air freshener has been reinvented, renamed and redesigned to rock out. Now known as Nest Audio, the new editions look more like a tiny, traditional speaker, this time in a multitude of colors (pink, blue, green, white and black), sell for less than the original Home ($99.99 versus $129.99) and the big news is a major sound upgrade. Nest Audio, available today, is still being sold as a personal assistant to run your smart home, answer trivia questions, set reminders, get news updates, translate languages and, of course, play music and podcasts. But the speaker, which runs on the Google Assistant, still lags Amazon's Echo speakers and the Alexa system in doing many obvious tasks that various Google help pages claims it can do, but either can't or require so much setup that consumers will be stymied. But let's start with what does work well: playing music. For $99, you get a speaker with vastly improved sound than the original, in a slightly larger body.

Understanding the importance of NoCode Machine learning tools and what the best ones in market? Businesses today run on information and insights, …

… to satellite data and machine learning, geospatial insights are now available to all. SkyWatch and Picterra are thrilled to announce their partnership …