scholarly journals Learning Actions That Reduce Variation in Objects

2021 ◽  
Author(s):  
◽  
James Bebbington

<p>The variation in the data that a robot in the real world receives from its sensory inputs (i.e. its sensory data) will come from many sources. Much of this variation is the result of ground truths about the world, such as what class an object belongs to, its shape, its condition, and so on. Robots would like to infer this information so they can use it to reason. A considerable amount of additional variation in the data, however, arises as a result of the robot’s relative configuration compared to an object; that is, its relative position, orientation, focal depth, etc. Fortunately, a robot has direct control over this configural variation: it can perform actions such as tilting its head or shifting its gaze. The task of inferring ground truth from data is difficult, and is made much more difficult when data is affected by configural variation. This thesis explores an approach in which the robot learns to perform actions that minimize the amount of configural variation in its sensory data, making the task of inferring information about objects considerably easier. The value of this approach is demonstrated by classifying digits from the MNIST and USPS datasets that have been transformed in various ways so that they include various kinds of configural variation.</p>

2021 ◽  
Author(s):  
◽  
James Bebbington

<p>The variation in the data that a robot in the real world receives from its sensory inputs (i.e. its sensory data) will come from many sources. Much of this variation is the result of ground truths about the world, such as what class an object belongs to, its shape, its condition, and so on. Robots would like to infer this information so they can use it to reason. A considerable amount of additional variation in the data, however, arises as a result of the robot’s relative configuration compared to an object; that is, its relative position, orientation, focal depth, etc. Fortunately, a robot has direct control over this configural variation: it can perform actions such as tilting its head or shifting its gaze. The task of inferring ground truth from data is difficult, and is made much more difficult when data is affected by configural variation. This thesis explores an approach in which the robot learns to perform actions that minimize the amount of configural variation in its sensory data, making the task of inferring information about objects considerably easier. The value of this approach is demonstrated by classifying digits from the MNIST and USPS datasets that have been transformed in various ways so that they include various kinds of configural variation.</p>


2019 ◽  
Vol 28 (3) ◽  
pp. 181-194
Author(s):  
Bernd Porr ◽  
Paul Miller

For an autonomous agent, the inputs are the sensory data that inform the agent of the state of the world, and the outputs are their actions, which act on the world and consequently produce new sensory inputs. The agent only knows of its own actions via their effect on future inputs; therefore desired states, and error signals, are most naturally defined in terms of the inputs. Most machine learning algorithms, however, operate in terms of desired outputs. For example, backpropagation takes target output values and propagates the corresponding error backwards through the network in order to change the weights. In closed loop settings, it is far more obvious how to define desired sensory inputs than desired actions, however. To train a deep network using errors defined in the input space would call for an algorithm that can propagate those errors forwards through the network, from input layer to output layer, in much the same way that activations are propagated. In this article, we present a novel learning algorithm which performs such ‘forward-propagation’ of errors. We demonstrate its performance, first in a simple line follower and then in a 1st person shooter game.


2010 ◽  
Vol 13 (1) ◽  
pp. 105-121
Author(s):  
Anik Waldow

This essay argues that Humean impressions are triggers of associative processes, which enable us to form stable patterns of thought that co-vary with our experiences of the world. It will thus challenge the importance of the Copy Principle by claiming that it is the regularity with which certain kinds of sensory inputs motivate certain sets of complex ideas that matters for the discrimination of ideas. This reading is conducive to Hume’s account of perception, because it avoids the impoverishment of conceptual resources so typical for empiricist theories of meaning and explains why ideas should be based on impressions, although impressions cannot be known to mirror matters of fact. Dieser Aufsatz argumentiert dafür, dass humesche Eindrücke („impressions“) Auslöser von assoziativen Prozessen sind, welche es uns ermöglichen, stabile Denkmuster zu bilden, die mit unseren Erfahrungen der Welt kovariant sind. Der Aufsatz stellt somit die Wichtigkeit des Kopien-Prinzips in Frage, nämlich dadurch, dass behauptet wird, für die Unterscheidung der Ideen sei die Regelmäßigkeit maßgeblich, mit der gewisse Arten von sensorischen Eingaben gewisse Mengen von komplexen Ideen motivieren. Diese Lesart trägt zu einem Verständnis von Humes Auffassung der Wahrnehmung bei, da sie die Verarmung der begrifflichen Mittel, die für empiristische Theorien der Bedeutung so typisch ist, vermeidet und erklärt, warum Ideen auf Eindrücken basieren sollten, obwohl Eindrücke nicht als Abbildungen von Tatsachen erkannt werden können.


2021 ◽  
Vol 13 (10) ◽  
pp. 5491
Author(s):  
Melissa Robson-Williams ◽  
Bruce Small ◽  
Roger Robson-Williams ◽  
Nick Kirk

The socio-environmental challenges the world faces are ‘swamps’: situations that are messy, complex, and uncertain. The aim of this paper is to help disciplinary scientists navigate these swamps. To achieve this, the paper evaluates an integrative framework designed for researching complex real-world problems, the Integration and Implementation Science (i2S) framework. As a pilot study, we examine seven inter and transdisciplinary agri-environmental case studies against the concepts presented in the i2S framework, and we hypothesise that considering concepts in the i2S framework during the planning and delivery of agri-environmental research will increase the usefulness of the research for next users. We found that for the types of complex, real-world research done in the case studies, increasing attention to the i2S dimensions correlated with increased usefulness for the end users. We conclude that using the i2S framework could provide handrails for researchers, to help them navigate the swamps when engaging with the complexity of socio-environmental problems.


Electronics ◽  
2021 ◽  
Vol 10 (8) ◽  
pp. 900
Author(s):  
Hanseob Kim ◽  
Taehyung Kim ◽  
Myungho Lee ◽  
Gerard Jounghyun Kim ◽  
Jae-In Hwang

Augmented reality (AR) scenes often inadvertently contain real world objects that are not relevant to the main AR content, such as arbitrary passersby on the street. We refer to these real-world objects as content-irrelevant real objects (CIROs). CIROs may distract users from focusing on the AR content and bring about perceptual issues (e.g., depth distortion or physicality conflict). In a prior work, we carried out a comparative experiment investigating the effects on user perception of the AR content by the degree of the visual diminishment of such a CIRO. Our findings revealed that the diminished representation had positive impacts on human perception, such as reducing the distraction and increasing the presence of the AR objects in the real environment. However, in that work, the ground truth test was staged with perfect and artifact-free diminishment. In this work, we applied an actual real-time object diminishment algorithm on the handheld AR platform, which cannot be completely artifact-free in practice, and evaluated its performance both objectively and subjectively. We found that the imperfect diminishment and visual artifacts can negatively affect the subjective user experience.


2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Sakthi Kumar Arul Prakash ◽  
Conrad Tucker

AbstractThis work investigates the ability to classify misinformation in online social media networks in a manner that avoids the need for ground truth labels. Rather than approach the classification problem as a task for humans or machine learning algorithms, this work leverages user–user and user–media (i.e.,media likes) interactions to infer the type of information (fake vs. authentic) being spread, without needing to know the actual details of the information itself. To study the inception and evolution of user–user and user–media interactions over time, we create an experimental platform that mimics the functionality of real-world social media networks. We develop a graphical model that considers the evolution of this network topology to model the uncertainty (entropy) propagation when fake and authentic media disseminates across the network. The creation of a real-world social media network enables a wide range of hypotheses to be tested pertaining to users, their interactions with other users, and with media content. The discovery that the entropy of user–user and user–media interactions approximate fake and authentic media likes, enables us to classify fake media in an unsupervised learning manner.


2020 ◽  
Vol 24 (09) ◽  

For the month of September 2020, APBN dives into the world of 3D printing and its wide range of real-world applications. Keeping our focus on the topic of the year, the COVID-19 pandemic, we explore the environmental impact of the global outbreak as well as gain insight to the top 5 vaccine platforms used in vaccine development. Discover more about technological advancements and how it is assisting innovation in geriatric health screening.


Target ◽  
2006 ◽  
Vol 18 (1) ◽  
pp. 17-47 ◽  
Author(s):  
Rainier Grutman

Texts foregrounding different languages pose unusual challenges for translators and translation scholars alike. This article seeks to provide some insights into what happens to multilingual literature in translation. First, Antoine Berman’s writings on translation are used to reframe questions of semantic loss in terms of the ideological underpinnings of translation as a cultural practice. This leads to a wider consideration of contextual aspects involved in the “refraction” of foreign languages, such as the translating literature’s relative position in the “World Republic of Letters” (Casanova). Drawing on a Canadian case-study (Marie-Claire Blais in English translation), it is suggested that asymmetrical relations between dominating and dominated literatures need not be negative per se, but can lead to the recognition of minority writers.


2005 ◽  
Vol 4 (3) ◽  
pp. 305-355
Author(s):  
Dušan Pokorný

AbstractThis chapter considers the meaning of the terms "society" and "market," and the need for markets to be institutionalized and legitimated. Obligatory norms and recommendatory guidelines today come from many sources: from states, from groupings of states, and from worldwide bodies such as the IMF, the WTO, and the World Bank. But when markets create profound inequalities both within and between societies, how do we determine what limits ought to be placed on markets? Since economic institutions are inseparable from culture, this is the "site" where the public will have to decide what is "society," what is the "market," and what will be the relation between them.


Sign in / Sign up

Export Citation Format

Share Document