Deep learning has been successfully applied to solve various complex problems ranging from big data analytics to computer vision and human-level control. Deep learning advances however have also been employed to create software that can cause threats to privacy, democracy and national security. One of those deep learning-powered applications recently emerged is deepfake. Deepfake algorithms can create fake images and videos that humans cannot distinguish them from authentic ones. The proposal of technologies that can automatically detect and assess the integrity of digital visual media is therefore indispensable. This paper presents a survey of algorithms used to create deepfakes and, more importantly, methods proposed to detect deepfakes in the literature to date. We present extensive discussions on challenges, research trends and directions related to deepfake technologies. By reviewing the background of deepfakes and state-of-the-art deepfake detection methods, this study provides a comprehensive overview of deepfake techniques and facilitates the development of new and more robust methods to deal with the increasingly challenging deepfakes.

Recently, blockchain-based federated learning (BFL) has attracted intensive research attention due to that the training process is auditable and the architecture is serverless avoiding the single point failure of the parameter server in vanilla federated learning (VFL). Nevertheless, BFL tremendously escalates the communication traffic volume because all local model updates (i.e., changes of model parameters) obtained by BFL clients will be transmitted to all miners for verification and to all clients for aggregation. In contrast, the parameter server and clients in VFL only retain aggregated model updates. Consequently, the huge communication traffic in BFL will inevitably impair the training efficiency and hinder the deployment of BFL in reality. To improve the practicality of BFL, we are among the first to propose a fast blockchain-based communication-efficient federated learning framework by compressing communications in BFL, called BCFL. Meanwhile, we derive the convergence rate of BCFL with non-convex loss. To maximize the final model accuracy, we further formulate the problem to minimize the training loss of the convergence rate subject to a limited training time with respect to the compression rate and the block generation rate, which is a bi-convex optimization problem and can be efficiently solved. To the end, to demonstrate the efficiency of BCFL, we carry out extensive experiments with standard CIFAR-10 and FEMNIST datasets. Our experimental results not only verify the correctness of our analysis, but also manifest that BCFL can remarkably reduce the communication traffic by 95-98% or shorten the training time by 90-95% compared with BFL.

K-means is one of the most widely used clustering models in practice. Due to the problem of data isolation and the requirement for high model performance, how to jointly build practical and secure K-means for multiple parties has become an important topic for many applications in the industry. Existing work on this is mainly of two types. The first type has efficiency advantages, but information leakage raises potential privacy risks. The second type is provable secure but is inefficient and even helpless for the large-scale data sparsity scenario. In this paper, we propose a new framework for efficient sparsity-aware K-means with three characteristics. First, our framework is divided into a data-independent offline phase and a much faster online phase, and the offline phase allows to pre-compute almost all cryptographic operations. Second, we take advantage of the vectorization techniques in both online and offline phases. Third, we adopt a sparse matrix multiplication for the data sparsity scenario to improve efficiency further. We conduct comprehensive experiments on three synthetic datasets and deploy our model in a real-world fraud detection task. Our experimental results show that, compared with the state-of-the-art solution, our model achieves competitive performance in terms of both running time and communication size, especially on sparse datasets.

Spoken language recognition (SLR) refers to the automatic process used to determine the language present in a speech sample. SLR is an important task in its own right, for example, as a tool to analyze or categorize large amounts of multi-lingual data. Further, it is also an essential tool for selecting downstream applications in a work flow, for example, to chose appropriate speech recognition or machine translation models. SLR systems are usually composed of two stages, one where an embedding representing the audio sample is extracted and a second one which computes the final scores for each language. In this work, we approach the SLR task as a detection problem and implement the second stage as a probabilistic linear discriminant analysis (PLDA) model. We show that discriminative training of the PLDA parameters gives large gains with respect to the usual generative training. Further, we propose a novel hierarchical approach where two PLDA models are trained, one to generate scores for clusters of highly-related languages and a second one to generate scores conditional to each cluster. The final language detection scores are computed as a combination of these two sets of scores. The complete model is trained discriminatively to optimize a cross-entropy objective. We show that this hierarchical approach consistently outperforms the non-hierarchical one for detection of highly related languages, in many cases by large margins. We train our systems on a collection of datasets including over 100 languages, and test them both on matched and mismatched conditions, showing that the gains are robust to condition mismatch.

The US Court of Appeals for the Federal Circuit affirmed on August 5 that only a natural person--not an artificial intelligence system--can be an inventor. Artificial Intelligence (AI) technology is widely applied as a tool in different technical areas, such as machine learning, image processing, and speech recognition. More complex AI technology can create new products or processes with little or no human help. If an AI system can independently create something new, can it be designated as an inventor? The Federal Circuit finally settled this issue--affirming decisions of the US Patent and Trademark Office (USPTO) and Eastern District of Virginia that an AI system cannot be an inventor.

Algorithmic Transparency in Action: How and Why Do Companies Disclose Information on Algorithms? Does the Accent of an Intelligent Conversational Agent Matter? What makes users reliant on the virtual assistant?

Today Insilico Medicine announced the completion of a second closing of its Series D round, led by Prosperity7 Ventures, the diversified growth fund of Saudi Aramco Ventures, bringing the total Series D financing to $95 million. Other global investors with expertise in the biopharmaceutical and life sciences sectors also participated. The financing brought in Prosperity7 as a new investor, alongside current investors in the Series D round, including a large, diversified asset management firm on the US West Coast, B Capital Group, Warburg Pincus, BHR Partners, Qiming Venture Partners, Deerfield, Pavilion Capital, BOLD Capital Partners, and WS Investment Company. Insilico's founder and CEO, Alex Zhavoronkov, PhD, also invested in the Series D round. Insilico Medicine plans to grow its presence in Saudi Arabia, building on the recent investment from Prosperity7.

A year ago researchers from the University of Zurich showcased their autonomous drones that were able to beat the fastest human pilots. However, that race wasn't "fair" in the sense that the AI algorithm commanding the drones had extra information that human pilots didn't have. In particular, the algorithm had access to near-perfect location and velocity estimation of the drones using motion capture systems, high-quality maps of the race course beforehand, and stereo cameras that can give depth information. This year, the team's autonomous drones raced on even playing fields without these handicaps, and its AI was able to beat the best human-controlled time by 0.5s in a three-lap race, a significant lead in the world of drone racing. Our take: This development is representative of AI progress ins many fields, where the researchers first make a working system with additional assumptions and then slowly chip away at these assumptions for a more robust and adaptable AI system.

Artificial Intelligence (AI) is a household name. All around us, businesses are increasingly leveraging AI for a wide range of uses. According to Allied Market Research, the global AI market size was valued at $65.48 billion in 2020, projected to reach $1,581.70 billion by 2030, growing at a CAGR of 38%. However, when it comes to women in AI, a World Economic Forum report states that only 22% of AI professionals globally are female, compared to 78% male, which accounts for a gender gap of 72%. And there is ample room for growth.

Semantic segmentation is a popular research topic in computer vision, and many efforts have been made on it with impressive results. In this paper, we intend to search an optimal network structure that can run in real-time for this problem. Towards this goal, we jointly search the depth, channel, dilation rate and feature spatial resolution, which results in a search space consisting of about 2.78*10^324 possible choices. To handle such a large search space, we leverage differential architecture search methods. However, the architecture parameters searched using existing differential methods need to be discretized, which causes the discretization gap between the architecture parameters found by the differential methods and their discretized version as the final solution for the architecture search. Hence, we relieve the problem of discretization gap from the innovative perspective of solution space regularization. Specifically, a novel Solution Space Regularization (SSR) loss is first proposed to effectively encourage the supernet to converge to its discrete one. Then, a new Hierarchical and Progressive Solution Space Shrinking method is presented to further achieve high efficiency of searching. In addition, we theoretically show that the optimization of SSR loss is equivalent to the L_0-norm regularization, which accounts for the improved search-evaluation gap. Comprehensive experiments show that the proposed search scheme can efficiently find an optimal network structure that yields an extremely fast speed (175 FPS) of segmentation with a small model size (1 M) while maintaining comparable accuracy.