A Continuous Semantic Embedding Method for Video Compact Representation

Video compact representation aims to obtain a representation that could reflect the kernel mode of video content and concisely describe the video. As most information in complex videos is either noisy or redundant, some researchers have instead focused on long-term video semantics. Recent video compact representation methods heavily rely on the segmentation accuracy of video semantics. In this paper, we propose a novel framework to address these challenges. Specifically, we designed a novel continuous video semantic embedding model to learn the actual distribution of video words. First, an embedding model based on the continuous bag of words method is proposed to learn the video embeddings, integrated with a well-designed discriminative negative sampling approach, which helps emphasize the convincing clips in the embedding while weakening the influence of the confusing ones. Second, an aggregated distribution pooling method is proposed to capture the semantic distribution of kernel modes in videos. Finally, our well-trained model can generate compact video representations by direct inference, which provides our model with a better generalization ability compared with those of previous methods. We performed extensive experiments on event detection and the mining of representative event parts. Experiments on TRECVID MED11 and CCV datasets demonstrated the effectiveness of our method. Our method could capture the semantic distribution of kernel modes in videos and shows powerful potential to discover and better describe complex video patterns.

A Joint Model Provisioning and Request Dispatch Solution for Low-Latency Inference Services on Edge

With the advancement of machine learning, a growing number of mobile users rely on machine learning inference for making time-sensitive and safety-critical decisions. Therefore, the demand for high-quality and low-latency inference services at the network edge has become the key to modern intelligent society. This paper proposes a novel solution that jointly provisions machine learning models and dispatches inference requests to reduce inference latency on edge nodes. Existing solutions either direct inference requests to the nearest edge node to save network latency or balance edge nodes’ workload by reducing queuing and computing time. The proposed solution provisions each edge node with the optimal number and type of inference instances under a holistic consideration of networking, computing, and memory resources. Mobile users can thus be directed to utilize inference services on the edge nodes that offer minimal serving latency. The proposed solution has been implemented using TensorFlow Serving and Kubernetes on an edge cluster. Through simulation and testbed experiments under various system settings, the evaluation results showed that the joint strategy could consistently achieve lower latency than simply searching for the best edge node to serve inference requests.

Application of first-order logic to identify organizers and perpetrators of illegal actions in teams of a limited circle of people

There are three main families of inference algorithms in first-order logic: direct inference and its application to deductive databases and production systems; backward inference procedures and logic programming systems; theorem proving systems based on the resolution method. When solving specific problems, the most effective algorithms are those that allow you to cover all the facts and axioms and must be taken into account in the process of inference. An example is considered in which it is necessary to prove the guilt of a person in murder. On the basis of statements, a knowledge base is formed from expressions, with the help of which an expression of first-order logic is compiled and proved using direct logical inference. The proof of the reasoning obtained in direct inference using the proof tree is given. However, direct inference provides for the implementation of all admissible stages of logical inference based on all known facts. The article also considers a method based on the resolution when implementing the reverse inference, taking into account the expression obtained in the direct inference. This expression is converted into a conjunctive normal formula using the laws of Boolean algebra and is proved by the elimination of events using the conjunction operation.

Politicizing the Environment

At the end of the 1950s, resource economists developed a method to derive demand functions for recreation sites from travel cost data for recreation planning purposes. Based on this work, a second, direct method of measurement was developed in the early sixties that became known as the contingent valuation method. Initially, this method asked respondents directly about their willingness to pay for a realistically described recreational amenity. When contingent valuation became used for valuation studies of environmental and health issues in a regulatory and legal framework, initial support for the method from resource and mainstream economists faded away, leading to a split in the profession between those who considered the method fit for this second purpose and those who considered this second use inappropriate and politically charged. Because much of this history has been told, including in this journal, the emphasis here is on the relation between indirect and direct inference pertaining to both methods, and the challenges that contingent valuation, as a method of direct inference, poses to the quality of a questionnaire and the possibilities of educating respondents in making a reasoned choice for the amenity on offer.

Tracking Cryptic SARS-CoV-2 Lineages Detected in NYC Wastewater

Tracking SARS-CoV-2 genetic diversity is strongly indicated because diversifying selection may lead to the emergence of novel variants resistant to naturally acquired or vaccine-induced immunity. To monitor New York City (NYC) for the presence of novel variants, we amplified regions of the SARS-CoV-2 Spike protein gene from RNA acquired from all 14 NYC wastewater treatment plants (WWTPs) and ascertained the diversity of lineages from these samples using high throughput sequencing. Here we report the detection and increasing frequencies of novel SARS-CoV-2 lineages not recognized in GISAIDs EpiCoV database. These lineages contain mutations rarely observed in clinical samples, including Q493K, Q498Y, H519N and T572N. Many of these mutations were found to expand the tropism of SARS-CoV-2 pseudoviruses by allowing infection of cells expressing the human, mouse, or rat ACE2 receptor. In addition, pseudoviruses containing the Spike amino acid sequence of these lineages were found to be resistant to many different classes of receptor binding domain (RBD) binding neutralizing monoclonal antibodies. We offer several hypotheses for the anomalous presence of these mutations, including the possibility of a non-human animal reservoir. Although wastewater sampling cannot provide direct inference of SARS-CoV-2 clinical sequences, our research revealed several lineages that could be relevant to public health and they would not have been discovered if not for wastewater surveillance.

Dynamic forecasting model for indoor pollutant concentration using recurrent neural network

Accurate and reliable indoor pollutant concentration prediction is essential to solve the time-lag problem of indoor air quality control systems. Thus, the representation of time in pollutant forecasting models is very important. One approach is to introduce an Elman neural network using a direct inference strategy into the time series forecast of indoor pollutant concentration. In this study, measurements of CO2 (ppm), total volatile organic compounds (mg/m3), particulate matter with a diameter smaller than 2.5 µm (PM2.5; µg/m3), the indoor dry bulb temperature (°C) and relative humidity (%) were carried out in a classroom at a middle school in Beijing, China. To identify air pollution antecedents, input selection was conducted based on correlation analysis. The results show that the information provided by the PM2.5 time series can better simulate the dynamic relationship between input and output data ([Formula: see text]= 0.963 and R2 = 0.928). In addition to the overall goodness of fit ([Formula: see text] = 0.982) of the CO2 time series, the peak and valley prediction capability of the model was evaluated using the relative peak error ( RPE) metric. Information from the valleys of the CO2 time series gives good results ([Formula: see text]). Therefore, a dynamic forecasting model with a direct inference strategy is a capable tool for identifying proper air pollution antecedents.

Direct Inference, Reichenbach's Principle, and the Sleeping Beauty Problem

A group of philosophers led by the late John Pollock has applied a method of reasoning about probability, known as direct inference and governed by a constraint known as Reichenbach's principle, to argue in support of ‘thirdism’ concerning the Sleeping Beauty Problem. A subsequent debate has ensued about whether their argument constitutes a legitimate application of direct inference. Here I defend the argument against two extant objections charging illegitimacy. One objection can be overcome via a natural and plausible definition, given here, of the binary relation ‘logically stronger than’ between two properties that can obtain even when the respective properties differ from one another in ‘arity’; given this definition, the Pollock group's argument conforms to Reichenbach's principle. Another objection prompts a certain refinement of Reichenbach's principle that is independently well-motivated. My defense of the Pollock group's argument has epistemological import beyond the Sleeping Beauty problem, because it both widens and sharpens the applicability of direct inference as a method for inferring single-case epistemic probabilities on the basis of general information of a probabilistic or statistical nature.