We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. By a carefully crafted design, we increased the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC14 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.

/pdf/going-deeper-with-convolutions-1yobw2o2ds.pdf

Going deeper with convolutions

Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

We will review some of the major results in random graphs and some of the more challenging open problems. We will cover algorithmic and structural questions. We will touch on newer models, including those related to the WWW.

Random graphs

There are a huge number of features which are said to improve Convolutional Neural Network (CNN) accuracy. Practical testing of combinations of such features on large datasets, and theoretical justification of the result, is required. Some features operate on certain models exclusively and for certain problems exclusively, or only for small-scale datasets; while some features, such as batch-normalization and residual-connections, are applicable to the majority of models, tasks, and datasets. We assume that such universal features include Weighted-Residual-Connections (WRC), Cross-Stage-Partial-connections (CSP), Cross mini-Batch Normalization (CmBN), Self-adversarial-training (SAT) and Mish-activation. We use new features: WRC, CSP, CmBN, SAT, Mish activation, Mosaic data augmentation, CmBN, DropBlock regularization, and CIoU loss, and combine some of them to achieve state-of-the-art results: 43.5% AP (65.7% AP50) for the MS COCO dataset at a realtime speed of ~65 FPS on Tesla V100. Source code is at this https URL

YOLOv4: Optimal Speed and Accuracy of Object Detection

“Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks”の学習報告

Metal halides (MHs) show great potential in photovoltaic and optoelectronic applications, but their lack of long-term stability limits their applications for their sensitivity to oxygen, moisture, strong light irradiation, and high temperature. In this report, lead-free MHs CsCu2I3/Cs3Cu2I5 dual-phase glass-ceramics were prepared with a traditional melting-quenching method by using solid reductive agent Si3N4. Ultrabroadband emission from violet to near-infrared (350–800 nm) can be obtained from these CsCu2I3/Cs3Cu2I5 dual-phase glass-ceramics with long-term water-resistance stability. Detailed analysis of microstructure, spectra, and calculation reveals that the ultrabroadband nature can be ascribed to the intense emission of CsCu2I3 NCs, Cu+ ions, and Cs3Cu2I5 NCs together, and especially, the emission from Cs3Cu2I5 NCs and Cu+ ions presents a cyan color. Our work not only demonstrates that CsCu2I3/Cs3Cu2I5 glass-ceramic is a promising fluorescent material, but also develops a method to prepare lead-free metal halides CsCu2I3/Cs3Cu2I5 dual-phase glass-ceramics with high stability, paving the way for their applications. 

Ultrabroadband Emission from CsCu2I3/Cs3Cu2I5 Dual-Phase Glass-Ceramics with Long-Term Stability

Generative Adversarial Networks (GAN) training process, in most cases, apply Uniform or Gaussian sampling methods in the latent space, which probably spends most of the computation on examples that can be properly handled and easy to generate. Theoretically, importance sampling speeds up stochastic optimization in supervised learning by prioritizing training examples. In this paper, we explore the possibility of adapting importance sampling into adversarial learning. We use importance sampling to replace Uniform and Gaussian sampling methods in the latent space and employ normalizing flow to approximate latent space posterior distribution by density estimation. Empirically, results on MNIST and Fashion-MNIST demonstrate that our method significantly accelerates GAN's optimization while retaining visual fidelity in generated samples.

FIS-GAN: GAN with Flow-based Importance Sampling

Location-based messaging applications (LMAs), a kind of messaging applications for mobile devices which enable users to connect with people based on their geographical locations, have recently experienced a huge popularity growth. The killer feature in LMAs that embodies the concept of geo-based instant messaging, named people nearby, allows users at any place to search and communicate with other registered users nearby. In this paper, we discuss a common weakness in LMAs that relates to the abuse of the people nearby function. In this case, rich personal data of registered LMA users can be easily obtained, bringing a chance to perform automated user profiling in LMAs. Specifically, we build an automated and scalable system to construct extended profiles (or we call life profile) of LMA users, which contain not only personal information of LMA users but also the daily activities and social ties inferred from their leaked spatio-temporal privacy. The system is highly adaptable to various applications, requiring no modification of applications or trivial work on protocol reverse engineering. We conduct the evaluation on a large scale for the first time. In our experiment, we succeed to construct life profiles for more than 280,000 users from two popular LMAs. The results of empirical analysis not only validate the existence of the privacy issue in LMAs, but also demonstrate its severity.

Automated User Profiling in Location-Based Mobile Messaging Applications

In one-shot weight sharing for NAS, the weights of each operation (at each layer) are supposed to be identical for all architectures (paths) in the supernet. However, this rules out the possibility of adjusting operation weights to cater for different paths, which limits the reliability of the evaluation results. In this paper, instead of counting on a single supernet, we introduce $K$-shot supernets and take their weights for each operation as a dictionary. The operation weight for each path is represented as a convex combination of items in a dictionary with a simplex code. This enables a matrix approximation of the stand-alone weight matrix with a higher rank ($K>1$). A \textit{simplex-net} is introduced to produce architecture-customized code for each path. As a result, all paths can adaptively learn how to share weights in the $K$-shot supernets and acquire corresponding weights for better evaluation. $K$-shot supernets and simplex-net can be iteratively trained, and we further extend the search to the channel dimension. Extensive experiments on benchmark datasets validate that K-shot NAS significantly improves the evaluation accuracy of paths and thus brings in impressive performance improvements.

Chang Xu

Papers

Ultrabroadband Emission from CsCu2I3/Cs3Cu2I5 Dual-Phase Glass-Ceramics with Long-Term Stability

FIS-GAN: GAN with Flow-based Importance Sampling

Automated User Profiling in Location-Based Mobile Messaging Applications

K-shot NAS: Learnable Weight-Sharing for NAS with K-shot Supernets

Combustion and gasification properties of petroleum coke and its pyrolytic semi-coke