Image captioning has increasingly large domains of application, and fashion is not an exception. Having automatic item descriptions is of great interest for fashion web platforms, sometimes hosting hundreds of thousands of images. This paper is one of the first to tackle image captioning for fashion images. To address dataset diversity issues, we introduced the InFashAIv1 dataset containing almost 16.000 African fashion item images with their titles, prices, and general descriptions. We also used the well-known DeepFashion dataset in addition to InFashAIv1. Captions are generated using the Show and Tell model made of CNN encoder and RNN Decoder. We showed that jointly training the model on both datasets improves captions quality for African style fashion images, suggesting a transfer learning from Western style data. The InFashAIv1 dataset is released on Github to encourage works with more diversity inclusion.

My name is Amena Khatun, and I currently live in Australia with my partner and our son. I am working as a'Postdoctoral Fellow' at Queensland University and Technology (QUT), Brisbane, Australia. My research interest is computer vision, deep learning, person re-identification, and security surveillance. In February 2017, my Ph.D. journey started in computer vision and deep learning at QUT. I am so thankful for the Australian Government RTP Scholarship, QUT HDR tuition Fees Sponsorship, and QUT Top-up Scholarship.

When a video purporting to show New Zealand Prime Minister Jacinda Ardern smoking drugs surfaced on social media in recent months, experts quickly dismissed it as a fake. The video, which was viewed and shared thousands of times, showed a woman smoking from what appeared to be a crack pipe. The PM's face had been superimposed using artificial intelligence. But the video, created for YouTube, was convincing enough to the many who shared it. It was the latest example of how disturbingly authentic-looking videos can blur the lines between reality and fantasy.

Building a socially intelligent agent involves many challenges, one of which is to teach the agent to speak guided by its value like a human. However, value-driven chatbots are still understudied in the area of dialogue systems. Most existing datasets focus on commonsense reasoning or social norm modeling. In this work, we present a new large-scale human value dataset called ValueNet, which contains human attitudes on 21,374 text scenarios. The dataset is organized in ten dimensions that conform to the basic human value theory in intercultural research. We further develop a Transformer-based value regression model on ValueNet to learn the utility distribution. Comprehensive empirical results show that the learned value model could benefit a wide range of dialogue tasks. For example, by teaching a generative agent with reinforcement learning and the rewards from the value model, our method attains state-of-the-art performance on the personalized dialog generation dataset: Persona-Chat. With values as additional features, existing emotion recognition models enable capturing rich human emotions in the context, which further improves the empathetic response generation performance in the EmpatheticDialogues dataset. To the best of our knowledge, ValueNet is the first large-scale text dataset for human value modeling, and we are the first one trying to incorporate a value model into emotionally intelligent dialogue systems. The dataset is available at https://liang-qiu.github.io/ValueNet/.

Bayesian neural networks (BNNs) have become a principal approach to alleviate overconfident predictions in deep learning, but they often suffer from scaling issues due to a large number of distribution parameters. In this paper, we discover that the first layer of a deep network possesses multiple disparate optima when solely retrained. This indicates a large posterior variance when the first layer is altered by a Bayesian layer, which motivates us to design a spatial-temporal-fusion BNN (STF-BNN) for efficiently scaling BNNs to large models: (1) first normally train a neural network from scratch to realize fast training; and (2) the first layer is converted to Bayesian and inferred by employing stochastic variational inference, while other layers are fixed. Compared to vanilla BNNs, our approach can greatly reduce the training time and the number of parameters, which contributes to scale BNNs efficiently. We further provide theoretical guarantees on the generalizability and the capability of mitigating overconfidence of STF-BNN. Comprehensive experiments demonstrate that STF-BNN (1) achieves the state-of-the-art performance on prediction and uncertainty quantification; (2) significantly improves adversarial robustness and privacy preservation; and (3) considerably reduces training time and memory costs.

Pinecone is a vector database for storing and searching through dense vectors. Why would you ever want to do that? Keep reading to find out. There is perhaps no greater contributor to the success of modern Natural Language Processing (NLP) technology than vector representations of language. The meteoric rise of NLP was ignited with the introduction of word2vec in 2013 [1]. Word2vec is one of the most iconic and earliest examples of dense vectors representing text.

It may be smart, but it's not that clever. Artificial intelligence is nothing without human input. The algorithms that drive AI rely on the expertise of programmers and it's still no more than a tool – albeit a powerful one – that scientists and engineers can use to solve problems. Yet this is not to say that AI isn't the fastest-growing deep technology in the world, with the potential to transform people's lives and boost nations' economies. Facilitating AI innovation has even become a priority for the UK government, as laid out in the National AI Strategy it published in September.

Lydia Winters shows off Microsoft's "Minecraft" ahead of the Electronic Entertainment Expo, in 2015. Cybersecurity experts say Minecraft players were quick to exploit a critical flaw in widely used software that intelligence firms raced to patch Friday. Lydia Winters shows off Microsoft's "Minecraft" ahead of the Electronic Entertainment Expo, in 2015. Cybersecurity experts say Minecraft players were quick to exploit a critical flaw in widely used software that intelligence firms raced to patch Friday. BOSTON -- A critical vulnerability in a widely used software tool -- one quickly exploited in the online game Minecraft -- is rapidly emerging as a major threat to organizations around the world.

Tiktoker Hannah Thelen and cyber journalist Kurt Knutsson join'Fox & Friends' to discuss what steps people can take to protect their data Security experts around the world raced Friday to patch one of the worst computer vulnerabilities discovered in years, a critical flaw in open-source code widely used across industry and government in cloud services and enterprise software. "I'd be hard-pressed to think of a company that's not at risk," said Joe Sullivan, chief security officer for Cloudflare, whose online infrastructure protects websites from malicious actors. Untold millions of servers have it installed, and experts said the fallout would not be known for several days. New Zealand's Computer emergency response team was among the first to report that the flaw in a Java-language utility for Apache servers used to log user activity was being "actively exploited in the wild" just hours after it was publicly reported Thursday and a patch released. The vulnerability, dubbed'Log4Shell,' was rated 10 on a scale of one to 10, the worst possible.

In recent years, there have been great advances in the field of decentralized learning with private data. Federated learning (FL) and split learning (SL) are two spearheads possessing their pros and cons, and are suited for many user clients and large models, respectively. To enjoy both benefits, hybrid approaches such as SplitFed have emerged of late, yet their fundamentals have still been illusive. In this work, we first identify the fundamental bottlenecks of SL, and thereby propose a scalable SL framework, coined SGLR. The server under SGLR broadcasts a common gradient averaged at the split-layer, emulating FL without any additional communication across clients as opposed to SplitFed. Meanwhile, SGLR splits the learning rate into its server-side and client-side rates, and separately adjusts them to support many clients in parallel. Simulation results corroborate that SGLR achieves higher accuracy than other baseline SL methods including SplitFed, which is even on par with FL consuming higher energy and communication costs. As a secondary result, we observe greater reduction in leakage of sensitive information via mutual information using SLGR over the baselines.