Assisted and automated driving functions will rely on machine learning algorithms, given their ability to cope with real-world variations, e.g. vehicles of different shapes, positions, colors, and so forth. Supervised learning needs annotated datasets, and several automotive datasets are available. However, these datasets are tremendous in volume, and labeling accuracy and quality can vary across different datasets and within dataset frames. Accurate and appropriate ground truth is especially important for automotive, as “incomplete” or “incorrect” learning can negatively impact vehicle safety when these neural networks are deployed. This work investigates the ground truth quality of widely adopted automotive datasets, including a detailed analysis of KITTI MoSeg. According to the identified and classified errors in the annotations of different automotive datasets, this article provides three different criteria collections for producing improved annotations. These criteria are enforceable and applicable to a wide variety of datasets. The three annotations sets are created to (i) remove dubious cases; (ii) annotate to the best of human visual system; and (iii) remove clear erroneous BBs. KITTI MoSeg has been reannotated three times according to the specified criteria, and three state-of-the-art deep neural network object detectors are used to evaluate them. The results clearly show that network performance is affected by ground truth variations, and removing clear errors is beneficial for predicting real-world objects only for some networks. The relabeled datasets still present some cases with “arbitrary”/“controversial” annotations, and therefore, this work concludes with some guidelines related to dataset annotation, metadata/sublabels, and specific automotive use cases.

Predictive maintenance attempts to prevent unscheduled downtime by scheduling maintenance before expected failures and/or breakdowns while maximally optimizing uptime. However, this is a non-trivial problem, which requires sufficient data analytics knowledge and labeled data, either to design supervised fault detection models or to evaluate the performance of unsupervised models. While today most companies collect data by adding sensors to their machinery, the majority of this data is unfortunately not labeled. Moreover, labeling requires expert knowledge and is very cumbersome. To solve this mismatch, we present an architecture that guides experts, only requiring them to label a very small subset of the data compared to today’s standard labeling campaigns that are used when designing predictive maintenance solutions. We use auto-encoders to highlight potential anomalies and clustering approaches to group these anomalies into (potential) failure types. The accompanied dashboard then presents the anomalies to domain experts for labeling. In this way, we enable domain experts to enrich routinely collected machine data with business intelligence via a user-friendly hybrid model, combining auto-encoder models with labeling steps and supervised models. Ultimately, the labeled failure data allows for creating better failure prediction models, which in turn enables more effective predictive maintenance. More specifically, our architecture gets rid of cumbersome labeling tasks, allowing companies to make maximum use of their data and expert knowledge to ultimately increase their profit. Using our methodology, we achieve a labeling gain of 90% at best compared to standard labeling tasks.

Work by Chomsky et al. (2019) and Epstein et al. (2018) develops a third-factor principle of computational efficiency called “Determinacy”, which rules out “ambiguous” syntactic rule-applications by requiring one-to-one correspondences between the input or output of a rule and a single term in the domain of that rule. This article first adopts the concept of “Input Determinacy” articulated by Goto and Ishii (2019, 2020), who apply Determinacy specifically to the input of operations like Merge, and then proposes to extend Determinacy to the labeling procedure developed by Chomsky (2013, 2015). In particular, Input Determinacy can explain restrictions on labeling in contexts where multiple potential labels are available (labeling ambiguity), and it can also provide an explanation for Chomsky's (2013, 2015) proposal that syntactic movement of an item (“Internal Merge”) renders that item invisible to the labeling procedure.

Discussions of terrorism assume actual or threatened violence, but the term is regularly used to delegitimize rivals' nonviolent actions. Yet do ordinary citizens accept descriptions of nonviolence as terrorism? Using a preregistered survey-experiment in Israel, a salient conflictual context with diverse repertoires of contention, we find that audiences rate adversary nonviolence close to terrorism, consider it illegitimate, and justify its forceful repression. These perceptions vary by the action's threatened harm, its salience, and respondents' ideology. Explicitly labeling nonviolence as terrorism, moreover, particularly sways middle-of-the-road centrists. These relationships replicate in a lower-salience conflict, albeit with milder absolute judgments, indicating generalizability. Hence, popular perceptions of terrorism are more fluid and manipulable than assumed, potentially undermining the positive effects associated with nonviolent campaigns.

In January 2020, the United States implemented a federal bioengineered labeling standard for food products that contain genetically modified material set to go into effect in January 2022. This bioengineered label indicates which products contain detectable levels of genetic material that have been modified through lab techniques that cannot be achieved in nature. An already existing alternative to the bioengineered label is the Non-GMO Project verified label which has been on the market since 2007, and indicates products free of genetically modified material through lab techniques. As consumers are now confronted with multiple labels pertaining to information related to genetic engineering, it is important to understand how people interpret these labels as it can lead to a greater understanding of how they inform consumer choice. We conducted a survey with 153 biology and environmental studies undergraduate students at Binghamton University in Binghamton, New York, asking questions about participants' views on genetically modified organisms (GMOs) and related terminology, corresponding food labels and how these labels influence their purchasing decisions. Results demonstrated a lack of awareness of the bioengineered label compared to the Non-GMO Project verified label. Additionally, individuals associated ‘bioengineered’ and ‘genetically modified’ with differing themes, where ‘bioengineered’ was more often associated with a scientific theme and ‘genetically modified’ was more often associated with an agricultural theme. There was also a discrepancy in how individuals said these labels influenced their purchases vs how the labels actually influenced purchasing decisions when participating in choice experiments. While the majority of participants reported that neither the Non-GMO Project verified label nor the bioengineered label influenced their purchasing decisions, in choice experiments, the majority of respondents chose products with the Non-GMO Project verified label. This study can give insight into overall perceptions of different terminologies associated with genetic engineering, in addition to how these labels are interpreted by consumers, and how they could affect purchasing decisions with the implementation of the new bioengineered label.

This article uses meta-regression analysis to examine variation in willingness to pay (WTP) for farm-raised seafood and aquaculture products. We measure the WTP premiums that consumers have for common product attributes and examine how WTP varies systematically across study design elements, populations of interest, and sample characteristics. Based on metadata from 45 studies, the meta-regression analysis indicates that WTP estimates differ significantly with the availability of attributes such as domestic and environmental certification, but also with sample income and gender representation.

The First Amendment to the US Constitution protects commercial speech from government interference. Commercial speech has been defined by the US Supreme Court as speech that proposes a commercial transaction, such as marketing and labeling. Companies that produce products associated with public health harms, such as alcohol, tobacco, and food, thus have a constitutional right to market these products to consumers. This article will examine the evolution of US law related to the protection of commercial speech, often at the expense of public health. It will then identify outstanding questions related to the commercial speech doctrine and the few remaining avenues available in the United States to regulate commercial speech including the use of government speech and addressing deceptive and misleading commercial speech.

Partial equilibrium models have been used extensively by policy makers to prospectively determine the consequences of government programs that affect consumer incomes or the prices consumers pay. However, these models have not previously been used to analyze government programs that inform consumers. In this paper, we develop a model that policy makers can use to quantitatively predict how consumers will respond to risk communications that contain new health information. The model combines Bayesian learning with the utility-maximization of consumer choice. We discuss how this model can be used to evaluate information policies; we then test the model by simulating the impacts of the North Dakota Folic Acid Educational Campaign as a validation exercise.

Understanding biofilm interactions with surrounding substratum and pollutants/particles can benefit from the application of existing microscopy tools. Using the example of biofilm interactions with zero-valent iron nanoparticles (nZVI), this study aims to apply various approaches in biofilm preparation and labeling for fluorescent or electron microscopy and energy dispersive X-ray spectrometry (EDS) microanalysis for accurate observations. According to the targeted microscopy method, biofilms were sampled as flocs or attached biofilm, submitted to labeling using 4’,6-diamidino-2-phenylindol, lectins PNA and ConA coupled to fluorescent dye or gold nanoparticles, and prepared for observation (fixation, cross-section, freezing, ultramicrotomy). Fluorescent microscopy revealed that nZVI were embedded in the biofilm structure as aggregates but the resolution was insufficient to observe individual nZVI. Cryo-scanning electron microscopy (SEM) observations showed nZVI aggregates close to bacteria, but it was not possible to confirm direct interactions between nZVI and cell membranes. Scanning transmission electron microscopy in the SEM (STEM-in-SEM) showed that nZVI aggregates could enter the biofilm to a depth of 7–11 µm. Bacteria were surrounded by a ring of extracellular polymeric substances (EPS) preventing direct nZVI/membrane interactions. STEM/EDS mapping revealed a co-localization of nZVI aggregates with lectins suggesting a potential role of EPS in nZVI embedding. Thus, the combination of divergent microscopy approaches is a good approach to better understand and characterize biofilm/metal interactions.

This paper investigates the state of pork supply as to its neglect of developing innovations and mechanisms for delivering superior eating quality to consumers. We explore reasons behind pork supply chains’ predominant focus on mass production combined with traceability and food safety, while only little attention has been given to potentially lucrative niche markets focused on intrinsic quality cues. Using established analytical frameworks of hedonic pricing and transactions costs economics we discuss alternative strategies for the segregation and promotion of intrinsic sensory differentiated pork. Growing empirical evidence in the literature underpins the importance of eating experience in delivering utility to consumers and in stabilizing declining demand trends in major markets. Building on current consumer behavioral literature and organizational developments in meat supply chains in Europe and Australia, we critically discuss opportunities to overcome this supposedly suboptimal situation.

We administered an online choice experiment to a sample of U.S. raw-oyster consumers to identify factors influencing preferences for Gulf of Mexico oysters, determined the extent of preference heterogeneity, and estimated marginal willingness to pay for specific varieties and other key attributes. Results indicate significant preference heterogeneity among select varieties, with non-Gulf respondents estimated to require a price discount on Gulf oyster varieties on the order of $3–$6/half dozen. Gulf respondents were found to be less sensitive to oyster variety, and estimated to be willing to pay a price premium only for select Gulf varieties on the order of $0–$3/half dozen.