Generative modeling is one of the hottest topics in AI. It's now possible to teach a machine to excel at human endeavors such as painting, writing, and composing music. With this practical book, machine-learning engineers and data scientists will discover how to re-create some of the most impressive examples of generative deep learning models, such as variational autoencoders,generative adversarial networks (GANs), encoder-decoder models, and world models. Author David Foster demonstrates the inner workings of each technique, starting with the basics of deep learning before advancing to some of the most cutting-edge algorithms in the field. Through tips and tricks, you'll understand how to make your models learn more efficiently and become more creative.

Fast forward 50 years and the processing power of the Apollo Guidance Computer (AGC) that took those men to the surface of the moon is now in your pocket, multiple times over-- in fact, an iPhone 6 could be used to guide 120 million Apollo 11 spacecraft to the moon, all at the same time. This factoid doesn't really do justice to the brilliance of the AGC. Given Moore's law, you could pick anything computational and say that 50 years later, there will exist a machine that can run it 2²⁵ faster. It was people like Margaret Hamilton, the lead for the software team who coded the AGC, who chose not to see current hardware limitations as a barrier, but instead as a challenge. She used the resource available to her at the time to achieve the unthinkable.

In this paper, we propose a new first-order gradient-based algorithm to train deep neural networks. We first introduce the sign operation of stochastic gradients (as in sign-based methods, e.g., SIGN-SGD) into ADAM, which is called as signADAM. Moreover, in order to make the rate of fitting each feature closer, we define a confidence function to distinguish different components of gradients and apply it to our algorithm. It can generate more sparse gradients than existing algorithms do. We call this new algorithm signADAM++. In particular, both our algorithms are easy to implement and can speed up training of various deep neural networks. The motivation of signADAM++ is preferably learning features from the most different samples by updating large and useful gradients regardless of useless information in stochastic gradients. We also establish theoretical convergence guarantees for our algorithms. Empirical results on various datasets and models show that our algorithms yield much better performance than many state-of-the-art algorithms including SIGN-SGD, SIGNUM and ADAM. We also analyze the performance from multiple perspectives including the loss landscape and develop an adaptive method to further improve generalization. The source code is available at https://github.com/DongWanginxdu/signADAM-Learn-by-Confidence.

The pursuit of explaining and improving generalization in deep learning has elicited efforts both in regularization techniques as well as visualization techniques of the loss surface geometry. The latter is related to the intuition prevalent in the community that flatter local optima leads to lower generalization error. In this paper, we harness the state-of-the-art "filter normalization" technique of loss-surface visualization to qualitatively understand the consequences of using adversarial training data augmentation as the explicit regularization technique of choice. Much to our surprise, we discover that this oft deployed adversarial augmentation technique does not actually result in "flatter" loss-landscapes, which requires rethinking adversarial training generalization, and the relationship between generalization and loss landscapes geometries.

The recent success of Deep Neural Networks (DNNs) has revealed the significant capability of neuromorphic computing in many challenging applications. Although DNNs are derived from emulating biological neurons, there still exist doubts over whether or not DNNs are the final and best model to emulate the mechanism of human intelligence. In particular, there are two discrepancies between computational DNN models and the observed facts of biological neurons. First, human neurons are interconnected randomly, while DNNs need carefully-designed architectures to work properly. Second, human neurons usually have a long spiking latency (~100ms) which implies that not many layers can be involved in making a decision, while DNNs could have hundreds of layers to guarantee high accuracy. In this paper, we propose a new computational neuromorphic model, namely shallow unorganized neural networks (SUNNs), in contrast to DNNs. The proposed SUNNs differ from standard ANNs or DNNs in three fundamental aspects: 1) SUNNs are based on an adaptive neuron cell model, Smart Neurons, that allows each neuron to adaptively respond to its inputs rather than carrying out a fixed weighted-sum operation like the neuron model in ANNs/DNNs; 2) SUNNs cope with computational tasks using only shallow architectures; 3) SUNNs have a natural topology with random interconnections, as the human brain does, and as proposed by Turing's B-type unorganized machines. We implemented the proposed SUNN architecture and tested it on a number of unsupervised early stage visual perception tasks. Surprisingly, such shallow architectures achieved very good results in our experiments. The success of our new computational model makes it a working example of Turing's B-Type machine that can achieve comparable or better performance against the state-of-the-art algorithms.

Action detection and recognition tasks have been the target of much focus in the computer vision community due to their many applications, namely, security, robotics and recommendation systems. Recently, datasets like AVA, provide multi-person, multi-label, spatiotemporal action detection and recognition challenges. Being unable to discern which portions of the input to use for classification is a limitation of two-stream CNN approaches, once the vision task involves several people with several labels. We address this limitation and improve the state-of-the-art performance of two-stream CNNs. In this paper we present four contributions: our fovea attention filtering that highlights targets for classification without discarding background; a generalized binary loss function designed for the AVA dataset; miniAVA, a partition of AVA that maintains temporal continuity and class distribution with only one tenth of the dataset size; and ablation studies on alternative attention filters. Our method, using fovea attention filtering and our generalized binary loss, achieves a relative video mAP improvement of 20% over the two-stream baseline in AVA, and is competitive with the state-of-the-art in the UCF101-24. We also show a relative video mAP improvement of 12.6% when using our generalized binary loss over the standard sum-of-sigmoids.

The Internet of Things (IoT) extends the Internet connectivity into billions of IoT devices around the world, which collect and share information to reflect the status of physical world. The Autonomous Control System (ACS), on the other hand, performs control functions on the physical systems without external intervention over an extended period of time. The integration of IoT and ACS results in a new concept - autonomous IoT (AIoT). The sensors collect information on the system status, based on which intelligent agents in IoT devices as well as Edge/Fog/Cloud servers make control decisions for the actuators to react. In order to achieve autonomy, a promising method is for the intelligent agents to leverage the techniques in the field of artificial intelligence, especially reinforcement learning (RL) and deep reinforcement learning (DRL) for decision making. In this paper, we first provide comprehensive survey of the state-of-art research, and then propose a general model for the applications of RL/DRL in AIoT. Finally, the challenges and open issues for future research are identified.

Deep reinforcement learning (DRL) has achieved great success in various applications. However, recent studies show that machine learning models are vulnerable to adversarial attacks. DRL models have been attacked by adding perturbations to observations. While such observation based attack is only one aspect of potential attacks on DRL, other forms of attacks which are more practical require further analysis, such as manipulating environment dynamics. Therefore, we propose to understand the vulnerabilities of DRL from various perspectives and provide a thorough taxonomy of potential attacks. We conduct the first set of experiments on the unexplored parts within the taxonomy. In addition to current observation based attacks against DRL, we propose the first targeted attacks based on action space and environment dynamics. We also introduce the online sequential attacks based on temporal consistency information among frames. To better estimate gradient in black-box setting, we propose a sampling strategy and theoretically prove its efficiency and estimation error bound. We conduct extensive experiments to compare the effectiveness of different attacks with several baselines in various environments, including game playing, robotics control, and autonomous driving.

Graph convolutional networks(GCNs) have become the most popular approaches for graph data in these days because of their powerful ability to extract features from graph. GCNs approaches are divided into two categories, spectral-based and spatial-based. As the earliest convolutional networks for graph data, spectral-based GCNs have achieved impressive results in many graph related analytics tasks. However, spectral-based models cannot directly work on directed graphs. In this paper, we propose an improved spectral-based GCN for the directed graph by leveraging redefined Laplacians to improve its propagation model. Our approach can work directly on directed graph data in semi-supervised nodes classification tasks. Experiments on a number of directed graph datasets demonstrate that our approach outperforms the state-of-the-art methods.

The physics informed neural network (PINN) is evolving as a viable method to solve partial differential equations. In the recent past PINNs have been successfully tested and validated to find solutions to both linear and non-linear partial differential equations (PDEs). However, the literature lacks detailed investigation of PINNs in terms of their representation capability. In this work, we first test the original PINN method in terms of its capability to represent a complicated function. Further, to address the shortcomings of the PINN architecture, we propose a novel distributed PINN, named DPINN. We first perform a direct comparison of the proposed DPINN approach against PINN to solve a non-linear PDE (Burgers' equation). We show that DPINN not only yields a more accurate solution to the Burgers' equation, but it is found to be more data-efficient as well. At last, we employ our novel DPINN to two-dimensional steady-state Navier-Stokes equation, which is a system of non-linear PDEs. To the best of the authors' knowledge, this is the first such attempt to directly solve the Navier-Stokes equation using a physics informed neural network.